diff --git a/agents/assistant/src/index.ts b/agents/assistant/src/index.ts
index 773a825644..bc2e799063 100644
--- a/agents/assistant/src/index.ts
+++ b/agents/assistant/src/index.ts
@@ -9,13 +9,13 @@
 // platform's deploy machinery; the execution host materializes it at
 // runtime from the deploy alone.
 //
-// Tool-package pins (CL-5999, CL-5852): `@intx/agent`'s `defineAgent`
+// Tool-package pins: `@intx/agent`'s `defineAgent`
 // still does not accept a `toolPackagePins` field on its authoring-time
 // config — it is vendored, read-only source for this change — so the
 // agent below is built directly against `AgentDefinition`'s own type,
 // which already carries the field, matching
 // `workflows/collateral-generation`'s precedent. `@corbits/memory`
-// (CL-8186) is pinned so this deployment can search, add, and list the
+// is pinned so this deployment can search, add, and list the
 // tenant's firm memory (`memory_search`/`memory_add`/`memory_list`) — a
 // published dependency, not a workbench-built tool package, resolved
 // through its own `interchange.tools` surface the same way any other
diff --git a/agents/assistant/src/system-prompt.ts b/agents/assistant/src/system-prompt.ts
index 16f6064812..e8a6fb955a 100644
--- a/agents/assistant/src/system-prompt.ts
+++ b/agents/assistant/src/system-prompt.ts
@@ -1,7 +1,7 @@
 // Myra's pure prompt/pin data — no runtime import of `@intx/workflow` or
 // `@intx/agent` (only a type-only import below), so this module is safe
-// to bundle into a browser build (CL-8154's `apps/web` needs exactly
-// this: the content, not `buildAssistantWorkflow`'s `defineWorkflow`
+// to bundle into a browser build (`apps/web` needs exactly this: the
+// content, not `buildAssistantWorkflow`'s `defineWorkflow`
 // call, which pulls in `@intx/agent`'s Node-bound runtime — file
 // locking, in particular — that a browser bundle cannot resolve).
 // `index.ts`'s `buildAssistantWorkflow` imports these same constants;
@@ -18,7 +18,7 @@ export const ASSISTANT_STEP_ID = "assistant";
  * these: `@corbits/memory` is a normal published dependency, not a
  * workspace tool package the pin/registry resolver can name, so its
  * tool factories are attached directly on `toolFactories` in
- * `./index.ts` instead (CL-8186). `@corbits/capability-tools` lets Myra
+ * `./index.ts` instead. `@corbits/capability-tools` lets Myra
  * self-service a missing tool, skill, or model; the manager-tools
  * bundles give Myra real workbench-management capability — a specialist
  * agent she can create (each gets their own chat), connection
@@ -63,7 +63,7 @@ const ASSISTANT_WELCOME_CLAUSE =
   "allowed to.";
 
 /**
- * TRIAGE (CL-6350): how Myra decides, on every message, whether to
+ * TRIAGE: how Myra decides, on every message, whether to
  * answer directly or delegate — what to do when a job needs a
  * connection this workbench doesn't have yet, and the reverse case: she
  * answers by default and only steps back when a message @-mentions a
@@ -95,14 +95,14 @@ const ASSISTANT_TRIAGE_CLAUSE =
   "unless you are @-mentioned too or the sender asks you directly.";
 
 /**
- * TEAMMATE (CL-7469 folds in COORDINATOR): how Myra offers help without
+ * TEAMMATE (folds in COORDINATOR): how Myra offers help without
  * pushing it — folds in the skills-capture nudge rather than a separate
  * always-on clause. Myra is the coordinator of this bench's agents, not
  * just its resident assistant: she never builds a teammate by
  * hand-editing files or answering in its place, and never stands up a
  * second instance of an agent that already exists.
  *
- * The build-arc sentence (CL-5879) is the load-bearing one: it never
+ * The build-arc sentence is the load-bearing one: it never
  * waits for someone to name the mechanism ("make an agent", "set up a
  * routine"). An *outcome* — a sales motion, a content pipeline, a repo
  * to keep up — is enough on its own for Myra to work out the
@@ -145,7 +145,7 @@ const ASSISTANT_TEAMMATE_CLAUSE =
   "built and what it can now do once it's live.";
 
 /**
- * DISCOVERY (CL-6179): the specific interview-then-propose procedure a
+ * DISCOVERY: the specific interview-then-propose procedure a
  * stated outcome triggers, on top of the general "work out the team
  * yourself" doctrine in `ASSISTANT_TRIAGE_CLAUSE` above — short and
  * concrete rather than open-ended, so discovery never turns into a
diff --git a/agents/assistant/test/manager-tools-scenario.test.ts b/agents/assistant/test/manager-tools-scenario.test.ts
index cb963d8f0b..0fc7c6fc85 100644
--- a/agents/assistant/test/manager-tools-scenario.test.ts
+++ b/agents/assistant/test/manager-tools-scenario.test.ts
@@ -1,5 +1,4 @@
-// A scripted playback of the owner's canonical GTM scenario (CL-5879
-// manager-tools follow-up), exercised directly against the manager-tools
+// A scripted playback of the owner's canonical GTM scenario, exercised directly against the manager-tools
 // bundles' `run()` with a single fake `fetch` standing in for every
 // workflow-run-authenticated route those bundles call. This proves the
 // CONTRACT across packages — a tool call's request shape matches what
@@ -132,9 +131,9 @@ function createFakeHub() {
     }
 
     if (url.pathname === "/api/workflow-chat/participants/messages" && method === "POST" && body) {
-      // The connect-service card post `request_connection` (CL-6393)
-      // makes into the caller's own room. Recorded so the scenario can
-      // assert the card itself, not just the tool's advice text.
+      // The connect-service card post `request_connection` makes into
+      // the caller's own room. Recorded so the scenario can assert the
+      // card itself, not just the tool's advice text.
       const parts = body["parts"] as {
         kind: string;
         block: { type: string; data: Record<string, unknown> };
@@ -244,7 +243,7 @@ async function runScenario(hub: ReturnType<typeof createFakeHub>): Promise<void>
   // A connector name that is neither a fixed `CONNECTOR_REGISTRY`
   // entry, a curated MCP preset, nor a connected MCP server makes
   // `request_connection` tell the agent to keep helping with what it
-  // can do (CL-6393) — never "go add a server and report back".
+  // can do — never "go add a server and report back".
   const unknown = await connectionsBundle.run(
     call("req_unknown", REQUEST_CONNECTION_TOOL, { connector: "acmecrm" }),
     new AbortController().signal,
@@ -255,7 +254,7 @@ async function runScenario(hub: ReturnType<typeof createFakeHub>): Promise<void>
   expect(hub.postedCards).toHaveLength(0);
 
   // She puts a connect card in the room for each of the three (real)
-  // services (CL-6393) — the card is what the human clicks; the tool
+  // services — the card is what the human clicks; the tool
   // result only tells Myra to keep helping in the meantime.
   for (const connector of ["exa", "granola", "linear"]) {
     const result = await connectionsBundle.run(
diff --git a/apps/web/README.md b/apps/web/README.md
index 1c612f34c4..349f7e8952 100644
--- a/apps/web/README.md
+++ b/apps/web/README.md
@@ -57,7 +57,7 @@ Running routines in the contextual column are sourced today from
 | `/c`         | Workbench deep-link surface. On wide layouts the conversation opens in the right canvas; on compact layouts it fills the main pane. Legacy `/chat` links still resolve here. |
 | `/workflows` | Workflow runs executing across your benches.                                                                                                                                 |
 | `/library`   | The artifact gallery. See "Library" below.                                                                                                                                   |
-| `/settings`  | Account, bench membership, and (CL-5990) Agents/Skills. See "Agents" and "Skills" below.                                                                                     |
+| `/settings`  | Account, bench membership, and Agents/Skills. See "Agents" and "Skills" below.                                                                                               |
 
 Agents and Skills are Settings sections, not rail destinations — old
 `/agents[/:id]` and `/skills[/:id]` links still resolve, redirecting to
diff --git a/apps/web/src/agent-chat-launch.ts b/apps/web/src/agent-chat-launch.ts
index cafb0ebef1..74ec787595 100644
--- a/apps/web/src/agent-chat-launch.ts
+++ b/apps/web/src/agent-chat-launch.ts
@@ -2,9 +2,9 @@
 // fresh chat with it" — the same `POST /workbenches` call this app's every
 // create path uses. `CreateAgentPanel`'s Settings → Agents entry point
 // calls this on success so an explicitly-defined new agent never ends
-// nowhere, and `instant-agent-create.ts` — THE one creation verb
-// (CL-6138) — calls it against the account's default setup template.
-// Always creates (CL-6089) — never the `reuseExisting` land-hop path,
+// nowhere, and `instant-agent-create.ts` — THE one creation verb —
+// calls it against the account's default setup template.
+// Always creates — never the `reuseExisting` land-hop path,
 // which is `default-agent-workbench.ts`'s own call, not this one.
 
 import { createWorkbench } from "@corbits/chat-ui";
diff --git a/apps/web/src/agent-dm-launch.ts b/apps/web/src/agent-dm-launch.ts
index 7bc9fcec63..4d28fa0bda 100644
--- a/apps/web/src/agent-dm-launch.ts
+++ b/apps/web/src/agent-dm-launch.ts
@@ -1,5 +1,5 @@
 // The one path from an agent definition's id to "the person is in their
-// direct chat with it" (CL-6253) — the sidebar's agent rows are the one
+// direct chat with it" — the sidebar's agent rows are the one
 // caller. Mirrors `agent-chat-launch.ts`'s shape exactly, but through
 // `openAgentDm` (`kind: "chat"`, `reuseExisting: true`) rather than
 // `createWorkbench` directly: the first click mints the DM, every later
diff --git a/apps/web/src/agents-api.ts b/apps/web/src/agents-api.ts
index 7e402aef0f..b76f4446b7 100644
--- a/apps/web/src/agents-api.ts
+++ b/apps/web/src/agents-api.ts
@@ -119,7 +119,7 @@ export function listAgentInstances(tenantId: string): Promise<readonly AgentInst
  * native `GET /workflows/runs` listing's own predicate (`address IS NOT
  * NULL AND anchorRunId = id`), the same one `isTopLevelRun` uses. Used
  * wherever a page needs "real deployments only" — the Agent Directory.
- * Single-tenant (CL-8087 accepted loss): the deleted route expanded the
+ * Single-tenant (accepted loss): the deleted route expanded the
  * requested tenant to its whole descendant subtree via
  * `getDescendantTenants`; the native listing filters one tenant, so a
  * workspace parent sees only its own runs, not its child workbenches'.
@@ -135,7 +135,7 @@ export function listTopLevelRuns(tenantId: string): Promise<readonly AgentInstan
  * bare-array discovery route at `/models` (`ModelInfo[]`) — those are
  * different wire shapes. Disabled rows are filtered out here because the
  * catalog may retain them. Embedding-named models, Hugging Face Hub paths,
- * and bare `.gguf` names are also omitted (CL-6744) — this endpoint carries
+ * and bare `.gguf` names are also omitted — this endpoint carries
  * no offering capability lists, so the name-only
  * {@link isChatPickerModelName} gate is the available signal. */
 export function listCatalogModels(tenantId: string): Promise<readonly CatalogModel[]> {
@@ -160,7 +160,7 @@ export type AgentDefinitionDraft = typeof AgentDefinitionDraftResponse.infer.dra
  * Asks Myra to draft a starting system prompt (and optionally a
  * refined description, a model pick, and skills) from a name and a
  * plain-language purpose — the create-agent panel's "Create & chat"
- * flow (CL-6074). Hits `@corbits/agent-directory`'s
+ * flow. Hits `@corbits/agent-directory`'s
  * `POST .../planner/agent-definitions/draft`; never deploys anything
  * itself. A caller that gets a rejected promise here (Myra unavailable,
  * the draft timing out, an unparseable or out-of-inventory reply, or a
diff --git a/apps/web/src/agents-directory.ts b/apps/web/src/agents-directory.ts
index bb543cf1a8..d045112485 100644
--- a/apps/web/src/agents-directory.ts
+++ b/apps/web/src/agents-directory.ts
@@ -18,7 +18,7 @@ import {
 
 import type { AgentDefinition, AgentInstance } from "./agents-api";
 
-/** An `AgentDefinition` with its display name derived (CL-6413) — its own
+/** An `AgentDefinition` with its display name derived — its own
  * description when the definition was created with one, otherwise a
  * humanized reading of its immutable `name` slug. `name` itself stays the
  * slug throughout; nothing here mutates it. */
@@ -41,7 +41,7 @@ export function purposeAgentDefinitions(
  * (see `./agents-api.ts`), which excludes every non-top-level run
  * (workbench host, invited agent, task) server-side — the native
  * `GET /workflows/runs` listing's own predicate. That listing is
- * single-tenant (CL-8087 accepted loss: no descendant-subtree rollup),
+ * single-tenant (accepted loss: no descendant-subtree rollup),
  * so this still applies the shared name-based workbench-host filter as
  * defense in depth.
  */
diff --git a/apps/web/src/api.ts b/apps/web/src/api.ts
index f541f8d065..f7e4b97693 100644
--- a/apps/web/src/api.ts
+++ b/apps/web/src/api.ts
@@ -53,8 +53,6 @@ export const ArtifactRowSchema = ArtifactListItemSchema.merge(
     content: "string",
   }),
 );
-// GET /api/tenants/:id/artifacts/:artifactId wraps the row in an `artifact`
-// envelope key (`@corbits/artifacts`' `mountArtifacts`, CL-8188).
 export const ArtifactDetailSchema = type({
   artifact: ArtifactRowSchema,
 });
@@ -179,7 +177,7 @@ export function rejectApproval(
 }
 
 /**
- * Sandboxed HTML preview URL for a Library artifact (CL-5879) — the same
+ * Sandboxed HTML preview URL for a Library artifact — the same
  * path an `<iframe sandbox>` in the canvas or Library detail pane loads,
  * and the "Open in new tab" affordance's `href`. Server-side (`GET
  * .../artifacts/:id/preview` in `@corbits/artifacts-hub`) answers 415 for
diff --git a/apps/web/src/app-error-boundary.tsx b/apps/web/src/app-error-boundary.tsx
index 638f544a9d..6357eb76db 100644
--- a/apps/web/src/app-error-boundary.tsx
+++ b/apps/web/src/app-error-boundary.tsx
@@ -1,4 +1,4 @@
-// The last line of defence (CL-6381): a render error anywhere in the tree
+// The last line of defence: a render error anywhere in the tree
 // used to leave the reader staring at a blank white page. React only offers
 // this as a class component (no hook equivalent exists), so it's the one
 // class in an otherwise function-component codebase.
diff --git a/apps/web/src/app.css b/apps/web/src/app.css
index bd6ce7925a..a86dda60b4 100644
--- a/apps/web/src/app.css
+++ b/apps/web/src/app.css
@@ -93,8 +93,8 @@ select:disabled,
   height: 100dvh;
 }
 
-/* A quoteable reference id under the error boundary's failure message
-   (CL-6632) — small and muted, same treatment as `.onboarding-error-refid`. */
+/* A quoteable reference id under the error boundary's failure message —
+   small and muted, same treatment as `.onboarding-error-refid`. */
 .app-boot-frame-refid {
   font-size: 0.75rem;
   color: var(--muted-foreground);
@@ -304,10 +304,10 @@ select:disabled,
   flex-direction: column;
 }
 
-/* CL-6092: the guided-fix banner, above the stage's own top bar — quiet
+/* the guided-fix banner, above the stage's own top bar — quiet
    until there is something to act on, then a clear (not alarming) nudge
    toward the fix. */
-/* Mount/unmount collapse (CL-6092): a grid-rows 0fr/1fr toggle so the
+/* Mount/unmount collapse: a grid-rows 0fr/1fr toggle so the
    banner's entrance/exit eases the stage's height rather than snapping it
    open or shut. `theme.css` (react-ui) already collapses every
    transition duration under `prefers-reduced-motion`, so this needs no
@@ -357,9 +357,9 @@ select:disabled,
   color: var(--card-foreground);
 }
 
-/* The Plugins gallery's own "couldn't find that connection" notice
-   (CL-6092): a deep link from the shell banner that names a provider the
-   gallery has no card for. */
+/* The Plugins gallery's own "couldn't find that connection" notice: a deep
+   link from the shell banner that names a provider the gallery has no
+   card for. */
 .plugins-connect-notice {
   display: flex;
   align-items: center;
@@ -1153,7 +1153,7 @@ select:disabled,
   overflow: auto;
 }
 
-/* Co-viewer cursors (CL-5958): positioned in the pane's own fractional
+/* Co-viewer cursors: positioned in the pane's own fractional
    coordinate space (see `PresenceCursor` in canvas-column.tsx), never
    intercepting pointer events — they are a read-only presence signal, not
    an interactive layer. */
@@ -1200,7 +1200,7 @@ select:disabled,
    that like any other. */
 
 /* Sidebar header: the brand mark, small and quiet — the mark is the brand
-   moment now (CL-6089 dropped the orange-rail era), not the switcher this
+ moment now (dropped the orange-rail era), not the switcher this
    dock replaced. */
 .shell-sidebar-brand-mark {
   display: flex;
@@ -1613,7 +1613,7 @@ select:disabled,
   white-space: nowrap;
 }
 
-/* Persistent pin glyph on pinned sidebar rows (CL-6657) — pinning must be
+/* Persistent pin glyph on pinned sidebar rows — pinning must be
    visible in the list itself, not only as a toast / settings bit. */
 .shell-ch-pin {
   flex-shrink: 0;
@@ -2269,8 +2269,8 @@ select:disabled,
   color: var(--muted-foreground);
 }
 
-/* A quoteable reference id under a consumer-language failure message
-   (CL-6360) — small and muted, never the focal point of the card. */
+/* A quoteable reference id under a consumer-language failure message —
+   small and muted, never the focal point of the card. */
 .onboarding-error-refid {
   font-size: 0.75rem;
   color: var(--muted-foreground);
@@ -2366,7 +2366,7 @@ select:disabled,
   min-height: 0;
 }
 
-/* Route-level Suspense fallback (CL-6370): centers the shared warm loader
+/* Route-level Suspense fallback: centers the shared warm loader
    instead of leaving a bare, un-laid-out box while a route chunk loads. */
 .shell-route-loading {
   display: flex;
@@ -3077,7 +3077,7 @@ tr.insights-row-clickable:hover {
   gap: 1rem;
 }
 
-/* CL-6124: the in-shell first-workbench screen — a chat, not a form. One
+/* the in-shell first-workbench screen — a chat, not a form. One
    prompt box, centered both ways in the stage; the stage owns the full
    width, no decorative panel. */
 .describe-first-workbench {
@@ -3124,7 +3124,7 @@ tr.insights-row-clickable:hover {
   color: var(--muted-foreground);
 }
 
-/* CL-6224: per-workbench Timeline — one wall-clock spine (chat messages,
+/* per-workbench Timeline — one wall-clock spine (chat messages,
    thread forks, routine runs, tasks, approvals), day dividers, and a right
    rail of per-day KPI counts. Matches the Insights stage's own visual
    language (hairline dividers, uppercase section labels, --shell-accent
@@ -3348,7 +3348,7 @@ tr.insights-row-clickable:hover {
   font-weight: 550;
 }
 
-/* ---------------- new-workbench picker (CL-6342) ---------------- */
+/* ---------------- new-workbench picker ---------------- */
 /* Screen 1 of the approved mock: one row list, no card grid. The
    selection rail (`box-shadow: inset 2px 0 0 0 var(--shell-accent)`)
    matches the sidebar's own active-row treatment exactly. */
@@ -3384,7 +3384,7 @@ tr.insights-row-clickable:hover {
   margin: 0;
 }
 
-/* The prompt box (CL-6628): the primary act on this screen, so it gets
+/* The prompt box: the primary act on this screen, so it gets
    the heaviest visual weight — a bordered, focus-ringed field with the
    send button inline, sized to read as "type here first". */
 .new-workbench-prompt {
@@ -3723,7 +3723,7 @@ tr.insights-row-clickable:hover {
   }
 }
 
-/* Mission Control (CL-6488/CL-6489) — stat strip, then a main column
+/* Mission Control — stat strip, then a main column
    (Needs you / In flight) beside a 320px rail (Jump back in / This week),
    the same "stack under 1100px" rule DESIGN.md gives every right rail. */
 .mission-control-layout {
diff --git a/apps/web/src/auto-workbench-title.test.ts b/apps/web/src/auto-workbench-title.test.ts
index 87abf2774b..ef3396ff58 100644
--- a/apps/web/src/auto-workbench-title.test.ts
+++ b/apps/web/src/auto-workbench-title.test.ts
@@ -7,7 +7,7 @@ import {
   titleFromFirstMessage,
 } from "./auto-workbench-title";
 
-describe("titleFromFirstMessage (CL-6656)", () => {
+describe("titleFromFirstMessage", () => {
   test("trims and collapses whitespace into a single-line title", () => {
     expect(titleFromFirstMessage("  Help me\nplan  Q3  ")).toBe("Help me plan Q3");
   });
@@ -33,7 +33,7 @@ describe("titleFromFirstMessage (CL-6656)", () => {
   });
 });
 
-describe("autoNameFromFirstMessage (CL-6656)", () => {
+describe("autoNameFromFirstMessage", () => {
   test("names an ad-hoc New Workbench from the first message", () => {
     expect(autoNameFromFirstMessage(NEW_WORKBENCH_TITLE, "Plan the Q3 launch")).toBe(
       "Plan the Q3 launch",
diff --git a/apps/web/src/auto-workbench-title.ts b/apps/web/src/auto-workbench-title.ts
index d13bda47c5..fed5ef361e 100644
--- a/apps/web/src/auto-workbench-title.ts
+++ b/apps/web/src/auto-workbench-title.ts
@@ -1,4 +1,4 @@
-// CL-6656: ad-hoc ("Just start talking") benches mint as "New Workbench" and
+// ad-hoc ("Just start talking") benches mint as "New Workbench" and
 // used to stay that way in the sidebar. Prefab templates already name the
 // room after the template title; this helper turns the person's first message
 // into a short sidebar title so blank benches get the same treatment once
diff --git a/apps/web/src/chat-artifact-open.ts b/apps/web/src/chat-artifact-open.ts
index a0ee1a75f8..33805b1455 100644
--- a/apps/web/src/chat-artifact-open.ts
+++ b/apps/web/src/chat-artifact-open.ts
@@ -29,7 +29,7 @@ export function artifactContentFromDetail(
     title: detail.title,
     rendererKind,
     content: detail.content,
-    // Text-kind Library artifacts are co-editable (CL-5958 phase 2); the
+    // Text-kind Library artifacts are co-editable (phase 2); the
     // presence `/update` route's own write-grant check is the real gate
     // — this only decides which pane a capable viewer sees.
     canEdit: rendererKind === "doc",
diff --git a/apps/web/src/client-bootstrap.ts b/apps/web/src/client-bootstrap.ts
index 85a48e8271..8064d5b625 100644
--- a/apps/web/src/client-bootstrap.ts
+++ b/apps/web/src/client-bootstrap.ts
@@ -44,7 +44,7 @@ function primaryTenantSlug(account: ClientBootstrapAccount): string {
   return base !== undefined && base.length > 0 ? base : `home-${account.id}`;
 }
 
-/** First-signup installer step (CL-8131): mints the account's primary
+/** First-signup installer step: mints the account's primary
  * tenant over the stock `POST /api/tenants` route (no `parentId`, so the
  * caller becomes its owner) when it does not already own one. This is
  * the "0→1" ruling's execution — never hub boot, never a CLI — and it
diff --git a/apps/web/src/command-palette-actions.test.ts b/apps/web/src/command-palette-actions.test.ts
index 38ab1d269e..b2d601914b 100644
--- a/apps/web/src/command-palette-actions.test.ts
+++ b/apps/web/src/command-palette-actions.test.ts
@@ -64,7 +64,7 @@ describe("ACTION_COMMANDS", () => {
     }
   });
 
-  test("exactly one New workbench create row — no duplicate title+destination (CL-6820)", () => {
+  test("exactly one New workbench create row — no duplicate title+destination", () => {
     const newWorkbenchRows = ACTION_COMMANDS.filter((c) => c.title === "New workbench");
     expect(newWorkbenchRows).toHaveLength(1);
     expect(newWorkbenchRows[0]?.id).toBe("new-workbench");
@@ -92,7 +92,7 @@ describe("ACTION_COMMANDS", () => {
 });
 
 describe("runActionCommand", () => {
-  test("new-workbench opens the template picker — no dialog, no pending flag, no instant mint (CL-6342)", async () => {
+  test("new-workbench opens the template picker — no dialog, no pending flag, no instant mint", async () => {
     const { ctx, navigated, dispatched } = context({ path: "/library" });
     await runActionCommand("new-workbench", ctx);
     expect(dispatched).toEqual([]);
diff --git a/apps/web/src/command-palette-actions.ts b/apps/web/src/command-palette-actions.ts
index 5b315b07a4..b45e22beab 100644
--- a/apps/web/src/command-palette-actions.ts
+++ b/apps/web/src/command-palette-actions.ts
@@ -6,16 +6,16 @@
 // a same-tick `dispatchEvent` would be a race the listener always loses.
 // `pending-dialog-request.ts` generalizes that pattern; the target page
 // (skills-page.tsx) consumes the pending flag on mount. Skills moved from
-// its own route into a Settings section (CL-5990) and back out to its own
-// rail destination (CL-6355) — "New skill" lands on `/skills`. "New task"
-// and the Inbox page it used to open are gone (CL-6151, owner decision):
+// its own route into a Settings section and back out to its own
+// rail destination — "New skill" lands on `/skills`. "New task"
+// and the Inbox page it used to open are gone (owner decision):
 // tasks are dispatched by Myra from inside a workbench now.
 //
 // Workbench creation is not one of those — there is no dialog to race, no
 // page to mount first: "new-workbench" navigates straight to the template
-// picker (`/new`, CL-6342 — superseding CL-6138's direct mint), the same
+// picker (`/new`, — superseding its direct mint), the same
 // hop the sidebar's own "+" control uses. A second "new-agent" row that
-// said and did the same thing was dropped (CL-6820). "New thread" is out
+// said and did the same thing was dropped. "New thread" is out
 // of scope (killed by owner decision).
 
 import { createPendingDialogRequest } from "@corbits/shell-layout";
diff --git a/apps/web/src/command-palette-provider.tsx b/apps/web/src/command-palette-provider.tsx
index d31887accb..c9f54d5afd 100644
--- a/apps/web/src/command-palette-provider.tsx
+++ b/apps/web/src/command-palette-provider.tsx
@@ -238,9 +238,9 @@ export function CommandPaletteProvider({
     open && selectedTenantId !== null,
     () => listSkills(selectedTenantId ?? ""),
   );
-  // Connected MCP servers for this bench. Until CL-6417 lands a real
+  // Connected MCP servers for this bench. Until lands a real
   // `/plugins/<slug>` page, selecting one opens the Plugins gallery
-  // (CL-6817) rather than a "still being built" stub.
+  // rather than a "still being built" stub.
   const mcpServersQuery = useTenantQuery(
     tenantKeys.mcpServers(selectedTenantId ?? ""),
     open && selectedTenantId !== null,
@@ -251,7 +251,7 @@ export function CommandPaletteProvider({
     ArtifactListPageSchema,
   );
 
-  // CL-6089's hidden escape hatch: the sidebar dropped its bench switcher
+  // its hidden escape hatch: the sidebar dropped its bench switcher
   // (a workbench IS a conversation now, one per account in the common
   // case), but a multi-bench install still needs a way in. Plainly
   // labeled, cycling to the next workbench in membership order — the
@@ -525,8 +525,8 @@ export function CommandPaletteProvider({
       } else if (id.startsWith("entity:plugins:")) {
         const slug = id.slice("entity:plugins:".length);
         const title = pluginItems.find((item) => item.id === id)?.title ?? slug;
-        // No plugin detail page yet (CL-6417 parked). Land on the gallery
-        // instead of the removed stub (CL-6817).
+        // No plugin detail page yet (parked). Land on the gallery
+        // instead of the removed stub.
         navigate(PLUGINS_PATH_PREFIX);
         pushRecent({ kind: "plugins", id, title, subtitle: "Plugin" });
       } else if (id.startsWith("entity:library:")) {
diff --git a/apps/web/src/connect-service-actions.ts b/apps/web/src/connect-service-actions.ts
index 01438a35dc..cc2dfc2cd6 100644
--- a/apps/web/src/connect-service-actions.ts
+++ b/apps/web/src/connect-service-actions.ts
@@ -1,6 +1,6 @@
 // Builds the `ConnectServiceActions` port `ChatWorkspace` (`@corbits/chat-ui`)
-// calls for the generic in-room connect card (CL-6393) — the room
-// connect-github binding's shape (hub-zero T3, CL-8114, deleted with the
+// calls for the generic in-room connect card — the room
+// connect-github binding's shape (hub-zero T3, deleted with the
 // workbench-scoped mount) generalized to every connector and
 // MCP preset. The block's data carries only agent-authored framing;
 // every live fact is resolved here: a curated preset reads the
@@ -14,7 +14,7 @@
 // credential completed without leaving the room (another tab, the
 // Plugins page) publishes `chat.settings`; ChatWorkspace parses that
 // event and calls `notifySettingsChanged` so a mounted card flips
-// without remounting (CL-6476).
+// without remounting.
 import type { ConnectServiceActions, ConnectServiceQuery } from "@corbits/chat-ui";
 import {
   completeConnectorCredential,
diff --git a/apps/web/src/granola-plugin-availability.ts b/apps/web/src/granola-plugin-availability.ts
index 5061d968b8..04d0ec8c75 100644
--- a/apps/web/src/granola-plugin-availability.ts
+++ b/apps/web/src/granola-plugin-availability.ts
@@ -1,6 +1,6 @@
 // Whether this tenant has Granola connected — the routine trigger popover
 // (see `shell/routine-panel.tsx`) reads this so an unconnected bench never
-// offers "Granola call notes" as if it were a working trigger (CL-6759).
+// offers "Granola call notes" as if it were a working trigger.
 // Tenant-scoped: Granola's credential resolves through the ancestor chain
 // (`GET /credentials/resolve/Granola`).
 import { type } from "arktype";
diff --git a/apps/web/src/insights-api.test.ts b/apps/web/src/insights-api.test.ts
index 924009e428..bb8aaf316c 100644
--- a/apps/web/src/insights-api.test.ts
+++ b/apps/web/src/insights-api.test.ts
@@ -3,9 +3,9 @@ import { describe, expect, test } from "bun:test";
 import { insightsTopLevelRunsPath } from "./insights-api";
 
 describe("insightsTopLevelRunsPath", () => {
-  // CL-8087: the native tenant-scoped top-level listing, not the deleted
+  // the native tenant-scoped top-level listing, not the deleted
   // `/top-level-runs` route (or its `feed=fires` variant) and not the dead
-  // `/me/workflows/runs`. CL-8160 deleted `packages/insights` and its
+  // `/me/workflows/runs`. deleted `packages/insights` and its
   // usage/activity/tools/latency/scope routes entirely.
   test("hits the native tenant-scoped GET /workflows/runs listing", () => {
     const path = insightsTopLevelRunsPath("tenant-1");
diff --git a/apps/web/src/insights-api.ts b/apps/web/src/insights-api.ts
index 5b70191b6f..11c6c2f248 100644
--- a/apps/web/src/insights-api.ts
+++ b/apps/web/src/insights-api.ts
@@ -1,13 +1,13 @@
 // This app's fetch/query plumbing over the native tenant-scoped
-// `GET /workflows/runs` top-level listing (CL-8087) →
-// Paginated<WorkflowRunResponse>. CL-8087 deleted `@corbits/run-scope`'s
+// `GET /workflows/runs` top-level listing →
+// Paginated<WorkflowRunResponse>. deleted `@corbits/run-scope`'s
 // `/top-level-runs?feed=fires`, which used to attribute each fire to its
-// routine server-side; CL-8160 deleted `packages/insights` itself — the
+// routine server-side; deleted `packages/insights` itself — the
 // hub mounts nothing Workbench-specific any more. The four stock
 // observability routes (`vendor/intx/hub-api/src/routes/observability.ts`)
 // are unimplemented stubs (each returns 501), so usage/cost/activity,
 // tool-call, latency, run-trace, and cross-workbench scope all lost their
-// only data source and are dropped rather than faked — see the CL-8160 PR
+// only data source and are dropped rather than faked — see the PR
 // description for the full accounting.
 
 import { type } from "arktype";
@@ -20,7 +20,7 @@ export const ListingTurnSchema = type({
 
 /**
  * `WorkflowRunResponse` plus the two routine-attribution fields the deleted
- * `feed=fires` mode of `/top-level-runs` used to report (CL-6249): the
+ * `feed=fires` mode of `/top-level-runs` used to report: the
  * routine that fired this run, when it has one. The native
  * `GET /workflows/runs` listing has no routine attribution, so both are
  * absent there — callers fall back to `definitionId`/`definitionName`
@@ -42,7 +42,7 @@ export const InsightsRunSchema = WorkflowRunResponse.and(
 );
 export type InsightsRun = typeof InsightsRunSchema.infer;
 
-/** Native `GET /workflows/runs` envelope (CL-8087). */
+/** Native `GET /workflows/runs` envelope. */
 export const TopLevelRunsSchema = paginatedSchema(InsightsRunSchema);
 
 // The REST pagination ceiling (see `vendor/intx/hub-api/src/pagination.ts`) —
@@ -50,13 +50,13 @@ export const TopLevelRunsSchema = paginatedSchema(InsightsRunSchema);
 const TOP_LEVEL_RUNS_LIMIT = 100;
 
 /**
- * Insights' run feed (CL-6062; repointed to native in CL-8087): the
+ * Insights' run feed (; repointed to native in): the
  * tenant's top-level runs from `GET /workflows/runs` — the native
  * listing's own predicate (`address IS NOT NULL AND anchorRunId = id`,
  * see `vendor/intx/hub-api/src/routes/runs.ts`) already excludes every
  * non-top-level run (workbench host, invited agent), so this page never
  * derives that exclusion itself. Four differences from the deleted
- * `feed=fires` feed, all recorded as accepted loss in CL-8087: a
+ * `feed=fires` feed, all recorded as accepted loss in: a
  * routine's fire (not a top-level run) is no longer included, so
  * Insights no longer sees routine executions; there is no routine
  * attribution, so history groups by definition; the resident,
@@ -65,7 +65,7 @@ const TOP_LEVEL_RUNS_LIMIT = 100;
  * hiding it; and the feed is single-tenant — the deleted route expanded
  * the requested tenant to its whole descendant subtree via
  * `getDescendantTenants` (the same rollup `resolveScope` used to give
- * `/usage`, `/activity`, and `/tools` before CL-8160 deleted them), while
+ * `/usage`, `/activity`, and `/tools` before deleted them), while
  * the native listing filters `tenantId = requested tenant` only, so a
  * workspace parent's runs feed no longer rolls up its child workbenches
  * and now sits mismatched against its own usage aggregate (moot now that
diff --git a/apps/web/src/insights-deeplinks.ts b/apps/web/src/insights-deeplinks.ts
index 143f586502..14c48e8fe1 100644
--- a/apps/web/src/insights-deeplinks.ts
+++ b/apps/web/src/insights-deeplinks.ts
@@ -15,7 +15,7 @@ export function runDeepLinkTarget(run: { readonly id: string }): string {
   return runDetailPath(run.id);
 }
 
-/** Insights scoped to one workbench (CL-5879) — the target for the
+/** Insights scoped to one workbench — the target for the
  * conversation action bar's "Insights" entry point. The route resolves the
  * workbench's own workbench tenant itself (see `../insights-workbench-scope.ts`)
  * rather than trusting a tenant id from the caller, since the caller here
diff --git a/apps/web/src/insights-path.ts b/apps/web/src/insights-path.ts
index 7d658188f2..18b9996135 100644
--- a/apps/web/src/insights-path.ts
+++ b/apps/web/src/insights-path.ts
@@ -7,7 +7,7 @@ import { decodedOrNull } from "@corbits/url-path";
 import { INSIGHTS_PATH_PREFIX, INSIGHTS_RUNS_PATH } from "./path-ids";
 
 /**
- * `/insights/workbench/:workbenchId` (CL-5879) is its own dedicated route — a
+ * `/insights/workbench/:workbenchId` is its own dedicated route — a
  * conversation's own scoped view, resolved by `InsightsWorkbenchPage`, never
  * a sub-mode of the landing. Every other path stays the cross-workbench
  * default landing: no per-mode branch needed there, since scoping happens
diff --git a/apps/web/src/insights-stats.test.ts b/apps/web/src/insights-stats.test.ts
index 7e426ade51..06764467c9 100644
--- a/apps/web/src/insights-stats.test.ts
+++ b/apps/web/src/insights-stats.test.ts
@@ -168,7 +168,7 @@ describe("groupRunsByDefinition", () => {
     expect(groupRunsByDefinition([])).toEqual([]);
   });
 
-  test("groups native rows with no routine attribution by definition (CL-8087)", () => {
+  test("groups native rows with no routine attribution by definition", () => {
     // The native `GET /workflows/runs` listing carries no `routineId`, so
     // history groups by definition — never by an invented routine.
     const native = run({
@@ -270,7 +270,7 @@ describe("runDisplayName", () => {
     ).toBe("researcher");
   });
 
-  test("falls back to the definition name for a native row with no routine attribution (CL-8087)", () => {
+  test("falls back to the definition name for a native row with no routine attribution", () => {
     const native = run({
       id: "native1",
       status: "running",
diff --git a/apps/web/src/insights-stats.ts b/apps/web/src/insights-stats.ts
index a5aeceaadb..c715e14b9e 100644
--- a/apps/web/src/insights-stats.ts
+++ b/apps/web/src/insights-stats.ts
@@ -8,7 +8,7 @@ import type { InsightsRun } from "./insights-api";
 import type { ScheduledWorkflowDefinition } from "./routines-api";
 
 /** Compact integer; null/undefined → em-dash. Lifted out of the deleted
- * `@corbits/insights/client` (CL-8160) — this app's own copy since it has
+ * `@corbits/insights/client` — this app's own copy since it has
  * no other browser-safe home now that the package is gone. */
 export function formatCount(value: number | null | undefined): string {
   if (value === null || value === undefined) return "—";
@@ -39,9 +39,9 @@ export const INSIGHTS_RECENT_LIMIT = 12;
 /**
  * Purpose runs only. `insights-page.tsx` sources `runs` from
  * `insightsTopLevelRunsPath` (see `./insights-api.ts`), the native `GET
- * /workflows/runs` top-level listing (CL-8087) whose own predicate already
+ * /workflows/runs` top-level listing whose own predicate already
  * excludes every non-top-level run — no workbench-host anchors exist to
- * filter post-cutover (CL-8175), so this is now an identity pass kept for
+ * filter post-cutover, so this is now an identity pass kept for
  * callers that still name it explicitly.
  */
 export function purposeRunsForInsights(runs: readonly InsightsRun[]): readonly InsightsRun[] {
@@ -49,10 +49,10 @@ export function purposeRunsForInsights(runs: readonly InsightsRun[]): readonly I
 }
 
 /**
- * A run's human-facing name (CL-6249): its routine's name when the feed
+ * A run's human-facing name: its routine's name when the feed
  * attributed one, honestly falling back to the definition name for a run
  * with no routine attribution — e.g. every row of the native
- * `GET /workflows/runs` listing (CL-8087), which carries no
+ * `GET /workflows/runs` listing, which carries no
  * `routineName`, or a directly launched workflow. Never mapped from a
  * client-side lookup table.
  */
@@ -68,7 +68,7 @@ export type DefinitionRunGroup = {
   /** `routineId` when the newest run in the group fired from one,
    * else `definitionId` — two different routines sharing one
    * definition (e.g. two workbench-digest schedules) never merge into
-   * one group. The native `GET /workflows/runs` feed (CL-8087) carries
+   * one group. The native `GET /workflows/runs` feed carries
    * no routine attribution, so in practice this is always `definitionId`
    * until a fires equivalent exists — the `routineId` branch is kept for
    * that feed, not removed. */
diff --git a/apps/web/src/insights-workbench-scope.ts b/apps/web/src/insights-workbench-scope.ts
index 461a534d76..d05cad4768 100644
--- a/apps/web/src/insights-workbench-scope.ts
+++ b/apps/web/src/insights-workbench-scope.ts
@@ -1,10 +1,10 @@
-// Insights is scoped per workbench (CL-5879): `/insights/workbench/:workbenchId`
+// Insights is scoped per workbench: `/insights/workbench/:workbenchId`
 // resolves to that workbench's OWN workbench tenant — every workbench minted
 // through POST /workbenches carries a tenancy link (see `WorkbenchWire.tenancy`
 // in @corbits/chat-ui) — never the workbench id itself and never the bench's
 // root tenant. That lookup goes through this one pure function over the SAME
 // cached workbench rows the shell's sidebar already fetches, rather than a
-// bespoke endpoint. CL-8160 deleted the reverse lookup
+// bespoke endpoint. deleted the reverse lookup
 // (`workbenchIdForWorkbenchTenant`): it only served the cross-workbench
 // "activity by workbench" chart and scope switcher, both dropped along with
 // packages/insights.
diff --git a/apps/web/src/instant-agent-create.test.ts b/apps/web/src/instant-agent-create.test.ts
index d9b1ccc677..6f141393f6 100644
--- a/apps/web/src/instant-agent-create.test.ts
+++ b/apps/web/src/instant-agent-create.test.ts
@@ -171,7 +171,7 @@ describe("createWorkbench", () => {
       if (path.endsWith("/chat/workbenches/chan-recoverable/invite")) {
         return json({ address: "def-assistant@tnt-1.corbits.dev", definitionId: "def-assistant" });
       }
-      // CL-8175: the opening message now posts through `@corbits/mailbox`'s
+      // the opening message now posts through `@corbits/mailbox`'s
       // `POST /me/inbox/send` rather than the chat route.
       if (path.endsWith("/mailbox/me/inbox/send")) {
         return json({ error: "agent launch failed" }, 409);
diff --git a/apps/web/src/instant-agent-create.ts b/apps/web/src/instant-agent-create.ts
index 9e7975b549..2895ef1ac0 100644
--- a/apps/web/src/instant-agent-create.ts
+++ b/apps/web/src/instant-agent-create.ts
@@ -1,9 +1,9 @@
 // Every "create a workbench" affordance — the sidebar's "+", the command
 // palette's "New workbench", and the zero-workbench land-hop on `/` —
 // opens the picker (`pages/new-workbench-picker.tsx`) and calls
-// `createWorkbench` below once a person hits Create. CL-8156 deleted
+// `createWorkbench` below once a person hits Create. deleted
 // workbench template picking (`@workbench/templates`): a new workbench is
-// a plain tenant + Myra now (CL-8154's converge path deploys Myra), so
+// a plain tenant + Myra now (its converge path deploys Myra), so
 // this mints an empty `kind: "workbench"` channel — no host, no
 // definitionId — and, when the prompt box supplied an opening message,
 // sends it and renames the room off it. Explicitly defining a brand-new
diff --git a/apps/web/src/last-workbench.ts b/apps/web/src/last-workbench.ts
index 6440abf650..4ec656e1a8 100644
--- a/apps/web/src/last-workbench.ts
+++ b/apps/web/src/last-workbench.ts
@@ -1,4 +1,4 @@
-// Files' workbench-first lens (CL-6353) needs to know which workbench, if
+// Files' workbench-first lens needs to know which workbench, if
 // any, the user was just inside — a signal `/files` can't carry itself,
 // since it's reached from the sidebar footer, not a workbench sub-route.
 // Recorded per bench (a switch to a different bench must not default to a
diff --git a/apps/web/src/library-workbench-scope.ts b/apps/web/src/library-workbench-scope.ts
index 6828728b7a..7af3dd003d 100644
--- a/apps/web/src/library-workbench-scope.ts
+++ b/apps/web/src/library-workbench-scope.ts
@@ -1,4 +1,4 @@
-// Files' workbench-first lens (CL-6353): resolves the "current workbench"
+// Files' workbench-first lens: resolves the "current workbench"
 // pill to that workbench's own tenant, the same `tenancy.tenantId`
 // `insights-workbench-scope.ts` scopes Insights on — never the workbench id
 // itself and never the bench's root tenant.
diff --git a/apps/web/src/login-next.test.ts b/apps/web/src/login-next.test.ts
index 645a643579..b55fdb01aa 100644
--- a/apps/web/src/login-next.test.ts
+++ b/apps/web/src/login-next.test.ts
@@ -1,4 +1,4 @@
-// CL-6369: the `next` param on /login is attacker-controllable (a crafted
+// the `next` param on /login is attacker-controllable (a crafted
 // `/login?next=...` link) — every case here is either a legitimate in-app
 // path or an open-redirect shape that must fall back to `/`.
 
diff --git a/apps/web/src/login-next.ts b/apps/web/src/login-next.ts
index d005abb80f..d7987dab61 100644
--- a/apps/web/src/login-next.ts
+++ b/apps/web/src/login-next.ts
@@ -1,4 +1,4 @@
-// CL-6369: URLs tell the truth. The `next` query param carries where a
+// URLs tell the truth. The `next` query param carries where a
 // signed-out visit was headed so a successful sign-in returns there — but
 // it's attacker-controllable (a crafted `/login?next=...` link), so it must
 // resolve to an in-app path or nowhere at all. Never an open redirect.
diff --git a/apps/web/src/main.tsx b/apps/web/src/main.tsx
index 61b0f08706..856d00aeb0 100644
--- a/apps/web/src/main.tsx
+++ b/apps/web/src/main.tsx
@@ -52,7 +52,7 @@ function Root() {
   // the hub's native setup-status route whether any bench exists yet. An
   // empty hub reports setup-required so we route into the setup screen;
   // a hub with tenants loads the shell normally. Read-only on purpose
-  // (CL-8112) — this never mints anything. A failure blocks the shell
+  // — this never mints anything. A failure blocks the shell
   // entirely rather than leaving the user silently benchless.
   const [provisioningError, setProvisioningError] = useState<{
     message: string;
@@ -91,7 +91,7 @@ function Root() {
 
   // Per-user storage when signed in so theme preference follows the account;
   // signed-out / loading share the anonymous host key. Not synced to the
-  // preferences store (CL-5922): @corbits/react-ui's ThemeProvider owns mode
+  // preferences store: @corbits/react-ui's ThemeProvider owns mode
   // entirely internally (localStorage read/write on setMode/cycleMode) and
   // exposes no onChange hook or externally-supplied initial value a host
   // could observe or override without forking the component.
diff --git a/apps/web/src/myra-deploy.ts b/apps/web/src/myra-deploy.ts
index 7d54e61648..aedb7efbe7 100644
--- a/apps/web/src/myra-deploy.ts
+++ b/apps/web/src/myra-deploy.ts
@@ -1,5 +1,5 @@
 // Builds and publishes Myra's deployable definition entirely over stock
-// routes (CL-8154). See `myra-source.ts` for why this is the one
+// routes. See `myra-source.ts` for why this is the one
 // deploy-source variant that installs today.
 //
 // Three steps, each a stock route: ensure the `package-registry` asset
diff --git a/apps/web/src/myra-source.ts b/apps/web/src/myra-source.ts
index 9a05230e33..a2e0ee9aa4 100644
--- a/apps/web/src/myra-source.ts
+++ b/apps/web/src/myra-source.ts
@@ -1,10 +1,10 @@
 // Myra's deploy source, pinned to what this hub can actually install
-// today (CL-8154).
+// today.
 //
 // `@corbits/assistant-workflow` (agents/assistant) is `private: true`
 // and unpublished, so `WorkflowDefinitionRegistrySource` (the `registry`
 // variant) has no external npm registry to resolve it from yet — that is
-// CL-8151's job. The package also exports builder functions
+// its job. The package also exports builder functions
 // (`buildAssistantWorkflow`), not a default-exported `WorkflowDefinition`,
 // so pointing a git `asset` "source" install at the real workspace
 // package would fail the source-tree installer's shape checks too.
diff --git a/apps/web/src/myra-workbench.ts b/apps/web/src/myra-workbench.ts
index 122792b5ed..9bb8db431f 100644
--- a/apps/web/src/myra-workbench.ts
+++ b/apps/web/src/myra-workbench.ts
@@ -26,7 +26,7 @@ const myraWorkbench = createDefaultAgentWorkbench({
 });
 
 /** The last workbench id `ensureMyraWorkbench` resolved to, for the shell's
- * col2-wide derivation (CL-5936): "Myra is the active surface" reduces to
+ * col2-wide derivation: "Myra is the active surface" reduces to
  * "the open workbench is the one Talk-to-Myra last landed us on". */
 export function isMyraWorkbenchId(workbenchId: string | null): boolean {
   return myraWorkbench.isCachedWorkbenchId(workbenchId);
diff --git a/apps/web/src/navigation.tsx b/apps/web/src/navigation.tsx
index fec5c0b8b8..c94f43bf61 100644
--- a/apps/web/src/navigation.tsx
+++ b/apps/web/src/navigation.tsx
@@ -24,7 +24,7 @@ const SignOutContext = createContext<(() => void) | undefined>(undefined);
 
 /** Same availability rule as `SignOutContext`: present in the signed-in
  * shell so surfaces like `ChatPage` can label the reader's own avatar from
- * the auth account (CL-6655), undefined outside that shell. */
+ * the auth account, undefined outside that shell. */
 const SessionUserContext = createContext<SessionUser | undefined>(undefined);
 
 export function NavigationProvider({
diff --git a/apps/web/src/needs-converge.ts b/apps/web/src/needs-converge.ts
index 206336d608..219b792a5e 100644
--- a/apps/web/src/needs-converge.ts
+++ b/apps/web/src/needs-converge.ts
@@ -99,7 +99,7 @@ export type StockHub = {
    * "member") in the same op — the stock invite route's `roleId` is the
    * only way an invited principal gets any grant at all; without it the
    * principal lands with zero roles and every subsequent read 403s
-   * (CL-8085/CL-8131 fix).
+   * (/ fix).
    */
   inviteMember(tenantId: string, input: { email: string; role: string }): Promise<void>;
   deployWorkflow(tenantId: string, input: WorkflowDeployInput): Promise<void>;
diff --git a/apps/web/src/onboarding.ts b/apps/web/src/onboarding.ts
index af3c993312..df40e2319c 100644
--- a/apps/web/src/onboarding.ts
+++ b/apps/web/src/onboarding.ts
@@ -1,8 +1,8 @@
 // The browser side of the first-login hook: reads the client converge
-// state (CL-8160) instead of a hub-served setup-status route — a hub with
+// state instead of a hub-served setup-status route — a hub with
 // zero owned tenants means setup-required so the UI routes into the setup
 // wizard; any owned tenant means a bench exists and the shell loads
-// normally. Read-only on purpose (CL-8112): provisioning itself back onto
+// normally. Read-only on purpose: provisioning itself back onto
 // the hub is T6/T7's surface, so this never mints anything — a failed
 // read blocks the shell loudly rather than leaving the user silently
 // benchless.
@@ -21,7 +21,7 @@ const CredentialsPage = type({
 /**
  * Outcome of the cheap credentials read used before trusting a
  * `seeded: true` hard-skip. A probe that cannot complete is never
- * collapsed into `none` (CL-6868): that would open paste-a-key as if no
+ * collapsed into `none`: that would open paste-a-key as if no
  * key exists when one may already be connected.
  */
 export type ActiveCredentialProbe =
@@ -33,7 +33,7 @@ export type ActiveCredentialProbe =
  * Whether `tenantId` has at least one credential in the `active`
  * status. A cheap, single read — no provider/catalog chain resolution —
  * so the caller treats it as the weaker check it is: only a confirmed
- * miss means "no key", never a probe failure (CL-6868).
+ * miss means "no key", never a probe failure.
  */
 export async function hasActiveCredential(tenantId: string): Promise<ActiveCredentialProbe> {
   try {
@@ -48,7 +48,7 @@ export async function hasActiveCredential(tenantId: string): Promise<ActiveCrede
   }
 }
 
-// The hub's user-facing error envelope (CL-6360): `userMessage` is
+// The hub's user-facing error envelope: `userMessage` is
 // consumer language, safe to render as-is; `refId` is what a person can
 // quote back for support. Never a raw `message`/stack/file-path field —
 // those stay in the hub's own logger.
@@ -84,7 +84,7 @@ export async function triggerFirstLoginProvisioning(): Promise<ProvisionOutcome>
   }
 }
 
-/* CL-8112 cut the rest of this module's provisioning surface: the
+/* cut the rest of this module's provisioning surface: the
  * credential-submit, one-click OAuth connect, and complete-setup helpers
  * all spoke to the deleted `/api/onboarding/*` routes, so they went with
  * the hub mount. What remains is the first-login status read above plus
@@ -93,7 +93,7 @@ export async function triggerFirstLoginProvisioning(): Promise<ProvisionOutcome>
  * native agent definition directly — so this helper exists only to keep
  * the deleted route's absence explicit: a 404/410 is `route-gone`, never
  * a generic agent error. The setup flow itself is T6/T7's to build.
- * TODO(CL-8112-T6/T7): replace this tombstone with native readiness once
+ * TODO(T6/T7): replace this tombstone with native readiness once
  * T6/T7 builds it — no new `/api/onboarding/*` client calls until then. */
 const ProvisioningStatus = type({
   kind: "'ready' | 'provisioning'",
@@ -102,7 +102,7 @@ const ProvisioningStatus = type({
 
 /**
  * Whether this account can start a conversation yet, and whether
- * anything is still coming online behind it (CL-6462). Deliberately not
+ * anything is still coming online behind it. Deliberately not
  * a count: how many workflows a bench seeds is an implementation detail,
  * and a person watching "0 of 5" learns nothing they can act on.
  *
@@ -112,7 +112,7 @@ const ProvisioningStatus = type({
  * - `preparing` — Myra is not live yet; this is the only state worth
  *   holding someone on a loader for.
  * - `route-gone` — the legacy `/api/onboarding/provisioning-status`
- *   route is gone (CL-8112 deleted it with the hub mount). Explicit on
+ * route is gone (deleted it with the hub mount). Explicit on
  *   purpose: a deleted route must never read as "your agent is broken".
  * - `error` — readiness could not be checked; show the message and allow retry.
  */
@@ -126,7 +126,7 @@ export type AgentReadiness =
 /**
  * Legacy agent-readiness read against the deleted
  * `/api/onboarding/provisioning-status` route. No live caller uses this
- * (CL-8112 T1): the home page's first-workbench flow and the picker read
+ * (T1): the home page's first-workbench flow and the picker read
  * Myra's native agent definition directly instead. Kept so the route's
  * absence stays an explicit, testable outcome — a 404/410 answers
  * `route-gone`, never the generic agent error below.
diff --git a/apps/web/src/onboarding/provider-connect-step.tsx b/apps/web/src/onboarding/provider-connect-step.tsx
index a7ff979527..ae95b6a1d1 100644
--- a/apps/web/src/onboarding/provider-connect-step.tsx
+++ b/apps/web/src/onboarding/provider-connect-step.tsx
@@ -1,5 +1,5 @@
 // The credential + model-offering step this onboarding page runs before
-// installing Myra (CL-8154): connects a provider credential through the
+// installing Myra: connects a provider credential through the
 // stock native credentials/providers routes (`@corbits/inference-settings`'
 // `shadowOffering`), then derives the exactly-one offering it mints into
 // `sourceOfferingIds`/`defaultSourceOfferingId` — no separate offering
@@ -28,7 +28,7 @@ export type ProviderOption = {
 // Deliberately small: `ModelProviderPlugin` names the adapters this hub
 // ships (`vendor/intx/types/src/catalog.ts`), and each option below picks
 // one canonical model so connecting a provider mints exactly one
-// offering — no second "which model" prompt (per the CL-8154 ruling,
+// offering — no second "which model" prompt (per the ruling,
 // the flow asks only where a human input is genuinely required).
 // Widening this list (more models per provider, an `openai-compatible`
 // custom-baseURL card) is a follow-up, not a blocker for a working
diff --git a/apps/web/src/pages/agent-detail-page.tsx b/apps/web/src/pages/agent-detail-page.tsx
index be1e17da06..e83db78e02 100644
--- a/apps/web/src/pages/agent-detail-page.tsx
+++ b/apps/web/src/pages/agent-detail-page.tsx
@@ -1,4 +1,4 @@
-// The agent's own page (CL-6414), addressed by its immutable slug —
+// The agent's own page, addressed by its immutable slug —
 // `/agents/<slug>`. Everything a person can author about an agent lives
 // here: its display name, the model it resolves against, the system prompt
 // it follows on every turn, the skills it has pinned, and the runs it has
diff --git a/apps/web/src/pages/agents-page.tsx b/apps/web/src/pages/agents-page.tsx
index c036497c1f..a809bb38d3 100644
--- a/apps/web/src/pages/agents-page.tsx
+++ b/apps/web/src/pages/agents-page.tsx
@@ -1,13 +1,12 @@
-// Agents: the global roster (CL-6354) — every agent definition this bench
+// Agents: the global roster — every agent definition this bench
 // has, one row per definition (name, description, workbenches currently
 // using it), with a detail panel that fetches the definition's model on
 // demand (`getAgentCapabilities`, the same route the per-workbench
 // Assistant editor reads). Create runs through `CreateAgentPanel`, the
-// same `createAgentDefinition` (`@corbits/agent-directory`) call the old
-// pre-CL-5990 Agents page used. Editing instructions/capabilities stays
-// where CL-6215 put it — the per-workbench Assistant tab
-// (`@corbits/chat-ui`'s `AgentsSection`) — this page is roster-only, never
-// a second instructions editor.
+// same `createAgentDefinition` (`@corbits/agent-directory`) call the
+// per-workbench Assistant tab uses. Editing instructions/capabilities
+// stays there (`@corbits/chat-ui`'s `AgentsSection`) — this page is
+// roster-only, never a second instructions editor.
 
 import {
   Badge,
@@ -166,7 +165,7 @@ export function archiveResultToast({ succeededIds, failedIds }: ArchiveDefinitio
 /** The short model name for a definition's capabilities — fetched lazily,
  * per row, the same route (and the same plain fetch-effect, no react-query
  * client required) `AgentDetailPanel` below already uses. A fetch failure
- * must not reuse the muted em-dash empty fields use (CL-6848). */
+ * must not reuse the muted em-dash empty fields use. */
 export type AgentModelCellState =
   | { readonly status: "loading" }
   | { readonly status: "ready"; readonly data: AgentCapabilities }
@@ -185,10 +184,10 @@ function catalogModelLabel(
 
 /** Settled Model-column content — an unset model reads as "Default"; a
  * fetch failure is a distinct error, never the same label. A set model
- * maps through the tenant catalog's displayName (CL-6748), the same
+ * maps through the tenant catalog's displayName, the same
  * `displayName ?? canonicalName` reading settings already uses — never
  * an invented label, and never a rewrite of an unset model to the
- * tenant default (CL-6782). */
+ * tenant default. */
 export function agentModelSettledContent(
   state:
     | { readonly status: "ready"; readonly data: AgentCapabilities }
@@ -273,7 +272,7 @@ function AgentModelCell({
 }
 
 /** Settled Runs · 7d content — a failed top-level-runs fetch is never the
- * same as an honest count of zero (CL-6842). */
+ * same as an honest count of zero. */
 export function agentRunsSettledContent(
   definitionId: string,
   instances: readonly AgentInstance[],
@@ -303,7 +302,7 @@ export type DefinitionWorkbenchInstance = {
  * definition id — the roster's "Workbenches" column and detail panel.
  * `chats` comes from `useBenchActivity`, the same agent-DM listing the
  * sidebar itself reads; the list here is exactly which rows the sidebar
- * would show for a definition before CL-6271's dedupe-by-title collapses
+ * would show for a definition before its dedupe-by-title collapses
  * same-named DMs across ancestor tenants. */
 export function workbenchesByDefinition(
   chats: readonly {
@@ -469,7 +468,7 @@ export function AgentsPage({
   readonly workbenches: ReadonlyMap<string, readonly DefinitionWorkbenchInstance[]>;
   readonly instances: readonly AgentInstance[];
   /** When the top-level-runs fetch failed — Status/Runs · 7d must not pretend
-   * the history is empty (CL-6842). */
+   * the history is empty. */
   readonly instancesError?: string | null;
   /** Tenant catalog used to map a stored canonical model id to its
    * person-readable displayName. Empty when the catalog failed independently
@@ -702,7 +701,7 @@ export function AgentsRoute({
   const directory = useAgentDirectory(selectedTenantId ?? undefined);
   const activity = useBenchActivity(selectedTenantId);
   // Powers the roster's Status and "Runs · 7d" columns. A failed fetch must
-  // not degrade those columns to Idle/0 (CL-6842) — the definitions listing
+  // not degrade those columns to Idle/0 — the definitions listing
   // still makes the page usable, but Status/Runs admit the load failed.
   const runsQuery = useQuery({
     queryKey: ["agent-top-level-runs", selectedTenantId],
diff --git a/apps/web/src/pages/chat-page.tsx b/apps/web/src/pages/chat-page.tsx
index 0d5e73fc01..d8f7447f0e 100644
--- a/apps/web/src/pages/chat-page.tsx
+++ b/apps/web/src/pages/chat-page.tsx
@@ -70,7 +70,7 @@ export function ChatPage({
   const queryClient = useQueryClient();
   const tenantId = bench.selectedTenantId;
 
-  // Same display name the sidebar account row already shows (CL-6655): the
+  // Same display name the sidebar account row already shows: the
   // auth session's name, which sign-up seeds from the email local-part when
   // no profile name was typed. Without this, chat-ui falls back to "Member"
   // for the reader's own presence/message avatar.
@@ -90,7 +90,7 @@ export function ChatPage({
             : {}),
         };
 
-  // Files' workbench-first lens (CL-6353) reads this back to default to
+  // Files' workbench-first lens reads this back to default to
   // "this workbench" when the person just came from one.
   useEffect(() => {
     if (tenantId === null || workbenchId === null) return;
@@ -98,7 +98,7 @@ export function ChatPage({
   }, [tenantId, workbenchId]);
 
   // Who's live in this workbench right now is derived inside `ChatWorkspace`
-  // itself now (CL-6328), off the same `/stream` connection as everything
+  // itself now, off the same `/stream` connection as everything
   // else — no separate `@corbits/presence` room/heartbeat for this surface
   // any more (that stack still backs the artifact canvas's cursor sync,
   // which has no chat stream of its own to piggyback on).
@@ -111,7 +111,7 @@ export function ChatPage({
     [tenantId, path],
   );
 
-  // The in-chat "Fix this connection" affordance's deep link (CL-6092) —
+  // The in-chat "Fix this connection" affordance's deep link —
   // the exact same hop the shell banner's own "Fix it" takes, reusing
   // `providerHealthBanner`'s current provider (chat-ui only sees a
   // classified reply's prose, never which provider it named — see
@@ -136,7 +136,7 @@ export function ChatPage({
     navigate("/plugins");
   }, [providerHealthBanner, requestPluginsConnect, navigate]);
 
-  // CL-6568: whether this tenant can actually run inference — never
+  // whether this tenant can actually run inference — never
   // whether a `model_provider` row merely exists, since seeding mints
   // that row with no credential attached. The same resolved-catalog
   // read `resolveModelSources` acts on at launch, so a model only
@@ -155,7 +155,7 @@ export function ChatPage({
   }, [navigate]);
 
   // A file part with an `artifactId` links back to a real Library row
-  // (CL-6000) — this always resolves through the Library artifacts read
+  // — this always resolves through the Library artifacts read
   // surface for that id, the same one `LibraryRoute` reads, never raw blob
   // bytes. Only a part with no `artifactId` (a plain human upload the
   // platform never diverted into an artifact) falls back to reading the
@@ -211,7 +211,7 @@ export function ChatPage({
     [navigate],
   );
 
-  // Hub-zero T3 (CL-8114): the room GitHub card is unbound on
+  // Hub-zero T3: the room GitHub card is unbound on
   // purpose — its state/start-reviewing routes are deleted and no native
   // equivalent exists yet, so it renders its no-port disabled framing
   // until a connections follow-up rebinds it. Connect/disconnect itself
diff --git a/apps/web/src/pages/create-agent-panel.tsx b/apps/web/src/pages/create-agent-panel.tsx
index bf9b04ec55..0c6857c767 100644
--- a/apps/web/src/pages/create-agent-panel.tsx
+++ b/apps/web/src/pages/create-agent-panel.tsx
@@ -1,25 +1,16 @@
-// The agent-create panel (CL-6074): the drafted create path for Settings →
-// Agents (`agents-page.tsx` mounts it). Agent configuration lives
-// per-workbench now (CL-6121); the "+ New workbench" picker used to open
-// this panel, then moved to instant creation (CL-6081) and, since CL-6138,
-// to THE one creation verb for that entry: `instant-agent-create.ts` mints
-// straight off the account's default setup template, no drafting, no
-// dialog. Agents still opens this panel when the person wants an explicit,
-// drafted agent — `describe-first-workbench.tsx` (CL-6104) once needed the
-// same drafting machinery — and on success it lands them in a chat with
-// the new agent (see `../agent-chat-launch.ts` for the shared "exact same
-// path picking an existing agent uses" hop).
+// The agent-create panel: the drafted create path for Settings → Agents.
+// On success it lands the person in a chat with the new agent (see
+// `../agent-chat-launch.ts` for the shared hop).
 //
 // Deliberately small above the fold: an identity swatch, a Name field,
 // and "Get started" is the whole happy path — a plain-language purpose
 // is optional, a quiet secondary field, not a gate. Either way, one
 // click on "Get started" asks Myra to draft a starting system prompt
 // (and optionally a description/model/skills) via `draftAgentDefinition`
-// (`@corbits/agent-directory`'s one-shot drafting port, CL-6074) — with a
-// purpose, from that brief; with none, a friendly general-purpose draft
-// the person teaches in the conversation that follows — then deploys
-// with that draft. The Suggestions row below the button is the same
-// flow with the name+purpose pre-written: every card describes
+// with a purpose, from that brief; with none, a friendly general-purpose
+// draft the person teaches in the conversation that follows — then
+// deploys with that draft. The Suggestions row below the button is the
+// same flow with the name+purpose pre-written: every card describes
 // something this workbench can actually do today. Handle, model
 // override, and skills sit behind "Advanced", collapsed by default, for
 // anyone who wants to steer them.
diff --git a/apps/web/src/pages/create-skill-dialog.tsx b/apps/web/src/pages/create-skill-dialog.tsx
index 1c1eda1dc2..b114d566c8 100644
--- a/apps/web/src/pages/create-skill-dialog.tsx
+++ b/apps/web/src/pages/create-skill-dialog.tsx
@@ -1,6 +1,6 @@
 // The create-skill dialog.
 //
-// CL-8086: this used to hand a full SKILL.md (name/description/body, or
+// this used to hand a full SKILL.md (name/description/body, or
 // an uploaded file) to the workbench's own skill registry, which parsed
 // and stored it. That registry is gone — skills are native `kind:"skill"`
 // hub assets now, and the stock asset routes (`@intx/hub-api`'s
diff --git a/apps/web/src/pages/home-page.tsx b/apps/web/src/pages/home-page.tsx
index 2005c8ebbd..fd5134f366 100644
--- a/apps/web/src/pages/home-page.tsx
+++ b/apps/web/src/pages/home-page.tsx
@@ -8,9 +8,8 @@
 // `/` only exists as this hop onto `/w/:workbenchId`. Deep links to
 // other pages are unchanged.
 //
-// Right after a provider connect this hop is also the wait (CL-6457's
-// deploys run in the background, so landing here can beat them). CL-6462
-// settled what that wait looks like: one warm loader and nothing else. For
+// Right after a provider connect this hop is also the wait (its
+// deploys run in the background, so landing here can beat them). // settled what that wait looks like: one warm loader and nothing else. For
 // a zero-workbench bench the wait is for Myra's own definition to exist at
 // all — then we open her DM. The check is simply retried every few seconds,
 // because Myra's readiness IS the test of whether the person can start —
@@ -20,7 +19,7 @@
 // genuine failure, never to draw a progress number: a seed count is an
 // implementation detail, and "0 of 5" told a waiting person nothing.
 //
-// CL-6780: that wait is for the agent, never a workbench that does not
+// that wait is for the agent, never a workbench that does not
 // exist yet — so the loader says "Preparing your agent", and a skip with
 // no credential stops pretending anything is "getting ready" and offers
 // the honest next step (connect a provider) instead of spinning forever.
@@ -48,7 +47,7 @@ type LandState =
   | { readonly kind: "opening" }
   /** Zero workbenches, credential present, Myra not ready yet — the
    * post-connect wait. Headline names the agent, never a workbench that
-   * does not exist (CL-6780). */
+   * does not exist. */
   | { readonly kind: "waiting-for-agent" }
   /** Myra has taken long enough that silence would read as a hang. Says
    * so plainly and offers another go — never a frozen number. */
@@ -103,14 +102,14 @@ export function HomeRoute({
 
     // Zero workbenches: wait for Myra's own definition to exist, then
     // open her DM the same way "Talk to Myra" does — never `/new`.
-    // Readiness is Myra's deployed definition itself (CL-8112 T1 cut the
+    // Readiness is Myra's deployed definition itself (T1 cut the
     // deleted `/api/onboarding/provisioning-status` read, so this flow
     // makes zero `/api/onboarding/*` requests): a definition means "she
     // can start", no definition means keep waiting. Without a credential
-    // the drain never starts (CL-6780), so no definition with no
+    // the drain never starts, so no definition with no
     // credential is an honest next step, not a forever spin on
     // "getting ready".
-    // TODO(CL-8112-T6/T7): read native readiness here once T6/T7 builds
+    // TODO(T6/T7): read native readiness here once T6/T7 builds
     // it, instead of the definition's existence.
     const awaitFirstWorkbench = () => {
       void listAgentDefinitions(selectedTenantId).then(
@@ -124,7 +123,7 @@ export function HomeRoute({
           void hasActiveCredential(selectedTenantId).then((probe) => {
             if (cancelled) return;
             // Only a confirmed miss means "connect a provider". A probe
-            // failure must not pretend no key exists (CL-6868) — keep
+            // failure must not pretend no key exists — keep
             // waiting and retry; the key may already be connected.
             if (probe.kind === "none") {
               setState({ kind: "needs-provider" });
diff --git a/apps/web/src/pages/insights-page-render.test.tsx b/apps/web/src/pages/insights-page-render.test.tsx
index 7d5c0d2ca8..5b8a184ab1 100644
--- a/apps/web/src/pages/insights-page-render.test.tsx
+++ b/apps/web/src/pages/insights-page-render.test.tsx
@@ -2,7 +2,7 @@
 // must render the same landing dashboard any other unrecognized Insights
 // path gets — never a blank page (see `insights-path.ts`'s
 // `parseInsightsPath`, which InsightsPage calls with the exact same `path`
-// prop this test passes). CL-8160 deleted the usage/activity/tools/latency/
+// prop this test passes). deleted the usage/activity/tools/latency/
 // scope-switcher props `InsightsPage` used to take along with
 // `packages/insights` itself — `InsightsWorkbenchPage`'s workbench-scoped
 // mode now lives entirely in `InsightsRoute`, so a `/insights/workbench/...`
diff --git a/apps/web/src/pages/insights-page.test.ts b/apps/web/src/pages/insights-page.test.ts
index d687cc896a..c1eac7a180 100644
--- a/apps/web/src/pages/insights-page.test.ts
+++ b/apps/web/src/pages/insights-page.test.ts
@@ -66,7 +66,7 @@ describe("isRunningNow", () => {
     ).toBe(false);
   });
 
-  // Warm-keep (CL-6681 / CL-6778): a routine's delivery agent stays deployed
+  // Warm-keep: a routine's delivery agent stays deployed
   // after it replies, so workflow_run.status lingers on `running`. Past the
   // fire window that is not an in-flight job — Insights must not keep it in
   // "Running now" forever.
diff --git a/apps/web/src/pages/insights-page.tsx b/apps/web/src/pages/insights-page.tsx
index bdf274be71..7882bce6e3 100644
--- a/apps/web/src/pages/insights-page.tsx
+++ b/apps/web/src/pages/insights-page.tsx
@@ -1,19 +1,11 @@
 // Insights over the native `GET /workflows/runs` top-level listing: KPI
 // row, "running now" strip, recent purpose runs, and runs history grouped
-// by definition. CL-8160: the hub mounts nothing Workbench-specific any
-// more (packages/insights and its `/insights/*` routes are gone), and the
-// stock observability routes it could have read instead
+// by definition. The stock observability routes
 // (`vendor/intx/hub-api/src/routes/observability.ts`'s four GET routes)
-// are unimplemented stubs — each returns 501. With no data source left,
-// this page dropped everything packages/insights used to compute: cost
-// KPIs, token mosaic, activity bars, cost-by-model, calls-by-tool, turn
-// latency, cross-workbench "activity by workbench" and the scope switcher
-// that navigated between sibling workbenches, and the run-detail trace
-// waterfall (steps/completed/failed/duration were derived from trace
-// spans this build no longer has). What is left — the run list, its KPI
-// counts, and run metadata — is entirely native `WorkflowRunResponse`
-// data, already fetched the same way `mission-control-page.tsx` and
-// `routines-api.ts` read runs elsewhere in this app.
+// are unimplemented stubs — each returns 501 — so this page is entirely
+// native `WorkflowRunResponse` data, already fetched the same way
+// `mission-control-page.tsx` and `routines-api.ts` read runs elsewhere in
+// this app.
 
 import {
   Badge,
@@ -628,7 +620,7 @@ export function InsightsPage({
 }
 
 /**
- * Insights scoped to one workbench (CL-5879) — `/insights/workbench/:workbenchId`
+ * Insights scoped to one workbench — `/insights/workbench/:workbenchId`
  * resolves the workbench's own workbench tenant (see
  * `../insights-workbench-scope.ts`) and titles the page by the WORKBENCH name,
  * never the tenant's. A true legacy workbench (tenancy `null`) and an id
diff --git a/apps/web/src/pages/legacy-settings-redirects.tsx b/apps/web/src/pages/legacy-settings-redirects.tsx
index 7715ca044f..0b54c920ad 100644
--- a/apps/web/src/pages/legacy-settings-redirects.tsx
+++ b/apps/web/src/pages/legacy-settings-redirects.tsx
@@ -1,9 +1,9 @@
 // Old links that must still land somewhere real. Agents and Skills were
-// Settings sections for a stretch (CL-5990: `/settings/agents[/:id]`,
-// `/settings/skills[/:id]`); CL-6354/CL-6355 moved both back out to their
+// Settings sections for a stretch (: `/settings/agents[/:id]`,
+// `/settings/skills[/:id]`); / moved both back out to their
 // own rail destinations, so any deep link into the old Settings home now
 // bounces to the new one, preserving a deep-linked id. Library was renamed
-// Files (CL-6353) at the same time it was moved off `/library` — that old
+// Files at the same time it was moved off `/library` — that old
 // prefix bounces to `/files` the same way.
 
 import { useEffect } from "react";
diff --git a/apps/web/src/pages/library-page.tsx b/apps/web/src/pages/library-page.tsx
index e8cd8abfc5..75fcd56434 100644
--- a/apps/web/src/pages/library-page.tsx
+++ b/apps/web/src/pages/library-page.tsx
@@ -174,7 +174,7 @@ function ArtifactRows({
 }
 
 /**
- * The one cheap provenance fact worth surfacing (CL-6015): a link to the
+ * The one cheap provenance fact worth surfacing: a link to the
  * workflow run that produced this artifact, when `source` says so
  * (`workflowRunIdFromSource`). Not a lineage system — every other origin
  * (manual, agent, imported, unknown) renders nothing here rather than
@@ -322,11 +322,11 @@ export function LibraryPage({
   readonly preview?: ArtifactDetail | null;
   readonly previewLoading?: boolean;
   readonly previewError?: string | null;
-  /** Needed to build the HTML preview route's URL (CL-5879); the "Open in
+  /** Needed to build the HTML preview route's URL; the "Open in
    * new tab" / iframe affordance is simply absent without one (a
    * standalone render with no bench tenant, e.g. these page tests). */
   readonly tenantId?: string | null;
-  /** The workbench the person just came from (CL-6353), if any — drives the
+  /** The workbench the person just came from, if any — drives the
    * "This workbench" pill. `null` when Files was reached with no workbench
    * in view, in which case the lens has nothing to offer and stays hidden. */
   readonly workbenchScope?: { readonly title: string } | null;
@@ -402,7 +402,7 @@ export function LibraryPage({
         subtitle={
           selectedSummary === null
             ? // Empty Files already has a poster invitation — a "0 files"
-              // count beside it is a second empty announcement (CL-6750).
+              // count beside it is a second empty announcement.
               artifacts.length === 0
               ? undefined
               : `${artifacts.length} files`
@@ -610,7 +610,7 @@ export function LibraryRoute({ path }: { readonly path: string }) {
   const [uploading, setUploading] = useState(false);
   const [uploadError, setUploadError] = useState<string | null>(null);
   const [searchQuery, setSearchQuery] = useState("");
-  // `/files/a/:id` (CL-6015) — a chat artifact chip's "Open in Files"
+  // `/files/a/:id` — a chat artifact chip's "Open in Files"
   // deep link, distinct from the kind-nav segments below. It only ever
   // sets the initial selection; the user's own clicks stay local state,
   // the same way kind-nav selection already worked before this route
@@ -622,7 +622,7 @@ export function LibraryRoute({ path }: { readonly path: string }) {
   }, [deepLinkedArtifactId]);
   const kindSegment = deepLinkedArtifactId === null ? libraryKindSegmentFromPath(path) : "";
 
-  // Files' workbench-first lens (CL-6353): the workbench the person just
+  // Files' workbench-first lens: the workbench the person just
   // came from, if `last-workbench.ts` recorded one for this bench, resolved
   // to its own tenant via the same sidebar-backed activity listing every
   // other bench-scoped surface already fetches.
diff --git a/apps/web/src/pages/mission-control-page.tsx b/apps/web/src/pages/mission-control-page.tsx
index 85245d15c7..1af30f79a3 100644
--- a/apps/web/src/pages/mission-control-page.tsx
+++ b/apps/web/src/pages/mission-control-page.tsx
@@ -1,11 +1,11 @@
-// Mission Control (CL-6488/CL-6489): the bench's dashboard — what needs a
+// Mission Control: the bench's dashboard — what needs a
 // decision, what's running, and a way back into recent context. A new
 // top-level route (`/mission-control`), never `/` — `/` stays the Myra
 // land-hop redirect (see routes.tsx's header comment). Every panel here is
 // backed by a query already used elsewhere in this app (pending approvals,
 // top-level runs); nothing on this page is invented. A panel with no honest
 // data source renders an empty state naming what's missing instead of a
-// fabricated number. CL-8160 dropped the "Runs today" / "Spend today" KPI
+// fabricated number. dropped the "Runs today" / "Spend today" KPI
 // tiles and the "This week" panel — both read packages/insights routes with
 // no stock equivalent.
 
@@ -317,7 +317,7 @@ export function MissionControlRoute({ navigate }: { readonly navigate: (to: stri
               inFlightRows.length === 0 ? (
                 <RichEmptyState
                   title="Nothing running right now"
-                  description="Routine activity has no feed in this build (CL-8087 deleted feed=fires with no native equivalent), so this stays empty until one lands."
+                  description="Routine activity has no feed in this build (deleted feed=fires with no native equivalent), so this stays empty until one lands."
                 />
               ) : null}
               {inFlightRows.length > 0 ? (
diff --git a/apps/web/src/pages/new-workbench-picker.tsx b/apps/web/src/pages/new-workbench-picker.tsx
index e22df41606..4d26eff589 100644
--- a/apps/web/src/pages/new-workbench-picker.tsx
+++ b/apps/web/src/pages/new-workbench-picker.tsx
@@ -1,5 +1,5 @@
-// CL-8156: workbench template picking is gone — a new workbench is a
-// plain tenant + Myra (CL-8154's converge path deploys Myra behind the
+// workbench template picking is gone — a new workbench is a
+// plain tenant + Myra (its converge path deploys Myra behind the
 // scenes). This screen is just the prompt box: say what you want and hit
 // Enter, or open an empty channel and add people/agents as you go.
 
@@ -258,7 +258,7 @@ export function NewWorkbenchPickerRoute() {
           // `delayMs={0}`: we already know this is a genuine wait the
           // instant the person hits Enter, so the default "hold back
           // briefly in case it resolves fast" delay only bought a blank
-          // pane here (CL-6623 finding #3) — show the loader outright
+          // pane here (finding #3) — show the loader outright
           // instead of leaving a gap before it mounts.
           <WorkbenchLoadingState delayMs={0} title="Setting up your workbench…" />
         ) : (
diff --git a/apps/web/src/pages/onboarding-page.tsx b/apps/web/src/pages/onboarding-page.tsx
index 75da7ec6ae..2be66fb256 100644
--- a/apps/web/src/pages/onboarding-page.tsx
+++ b/apps/web/src/pages/onboarding-page.tsx
@@ -1,4 +1,4 @@
-// The setup gate (CL-8112/CL-8131/CL-8154): the screen a signed-in
+// The setup gate: the screen a signed-in
 // session lands on when the hub reports setup-required. It reads the
 // hub's native setup-status route, and a hub that already has tenants
 // bounces straight into the shell (`/`). An empty hub drives the
diff --git a/apps/web/src/pages/plugins-page.tsx b/apps/web/src/pages/plugins-page.tsx
index 6702ae3b4c..403395a0cd 100644
--- a/apps/web/src/pages/plugins-page.tsx
+++ b/apps/web/src/pages/plugins-page.tsx
@@ -1,4 +1,4 @@
-// The Plugins gallery route (CL-6090) — a thin composition, same shape as
+// The Plugins gallery route — a thin composition, same shape as
 // `library-page.tsx`: page-level data fetching and wiring live here,
 // presentation lives in `@corbits/plugins-ui`. This page reads the
 // currently selected bench tenant the same way every other Settings
@@ -77,11 +77,11 @@ export function PluginsRoute({
   const [activeTab, setActiveTab] = useState<PluginsGalleryTab>("plugins");
   const [galleryQuery, setGalleryQuery] = useState("");
   // Set when the shell banner's "Fix it" deep link named a provider this
-  // gallery has no matching card for (CL-6092) — rather than the deep
+  // gallery has no matching card for — rather than the deep
   // link silently no-oping, this renders a notice pointing back at the
   // gallery itself.
   const [connectDeepLinkNotFound, setConnectDeepLinkNotFound] = useState(false);
-  // A preset deep link's slug (CL-7141), passed to the MCP presets
+  // A preset deep link's slug, passed to the MCP presets
   // section so it can focus that preset's own card once its catalog
   // has loaded — cleared as soon as the section has acted on it.
   const [autoConnectPresetSlug, setAutoConnectPresetSlug] = useState<string | null>(null);
@@ -182,7 +182,7 @@ export function PluginsRoute({
     };
   }, [selectedTenantId, skillsReloadKey]);
 
-  // The shell banner's "Fix it" deep link (CL-6092): once the gallery has
+  // The shell banner's "Fix it" deep link: once the gallery has
   // loaded, pick up any pending provider id and open its connect panel —
   // the same panel a gallery card click opens, so a person lands exactly
   // where they would have clicked themselves.
@@ -200,7 +200,7 @@ export function PluginsRoute({
     clearPendingConnectProvider();
   }, [pluginsState, pendingConnectProvider, clearPendingConnectProvider, openPluginPanel]);
 
-  // `request_connection`'s fallback link (CL-7141): `/plugins?connect=<id>`
+  // `request_connection`'s fallback link: `/plugins?connect=<id>`
   // hands the connector id off through the same `requestPluginsConnect`
   // path the shell banner's "Fix it" click uses, then strips only the
   // `connect` param — any other query param this route is ever opened
diff --git a/apps/web/src/pages/routines-available-strings.ts b/apps/web/src/pages/routines-available-strings.ts
index 031ecdaf9b..656882e68a 100644
--- a/apps/web/src/pages/routines-available-strings.ts
+++ b/apps/web/src/pages/routines-available-strings.ts
@@ -1,4 +1,4 @@
-// Copy for the Routines page's "Available" section (CL-7073) — kept apart
+// Copy for the Routines page's "Available" section — kept apart
 // from JSX so a wording change never touches component logic. Every
 // string here reads as consumer language: no "deploy", "asset", or other
 // platform-side noun (see `scripts/checks/ui-vocabulary.ts`).
diff --git a/apps/web/src/pages/routines-page.tsx b/apps/web/src/pages/routines-page.tsx
index a7117d79e0..0b1ca6b03c 100644
--- a/apps/web/src/pages/routines-page.tsx
+++ b/apps/web/src/pages/routines-page.tsx
@@ -1,7 +1,7 @@
 // Routines: an ops table of authored workflow definitions that carry a
 // ScheduleTrigger, including paused (`stopped`) ones. Pause/resume and
 // run-now are the only writes; schedules are authored on the definition.
-// The "Available" section (CL-7073) is scoped to the current bench only
+// The "Available" section is scoped to the current bench only
 // — adding a catalog workflow is a single-tenant write, unlike the
 // scheduled roster above, which aggregates every bench the account
 // belongs to.
diff --git a/apps/web/src/pages/run-status-tone-parity.test.ts b/apps/web/src/pages/run-status-tone-parity.test.ts
index 805870ebb3..b87b55dbfd 100644
--- a/apps/web/src/pages/run-status-tone-parity.test.ts
+++ b/apps/web/src/pages/run-status-tone-parity.test.ts
@@ -1,4 +1,4 @@
-// Guards against the exact bug this file was added for (CL-6499 design
+// Guards against the exact bug this file was added for (design
 // review): a page grows its own run-status → tone map/function instead of
 // reading react-ui's `RUN_STATUS_TONE` (`workflow-run.ts`), and it quietly
 // disagrees — a cancelled run reading neutral grey on Routines and amber
diff --git a/apps/web/src/pages/skill-detail-page.tsx b/apps/web/src/pages/skill-detail-page.tsx
index 1ed6fbccd7..618aba2d37 100644
--- a/apps/web/src/pages/skill-detail-page.tsx
+++ b/apps/web/src/pages/skill-detail-page.tsx
@@ -1,6 +1,6 @@
-// The skill detail page at `/skills/<name>` (CL-6416).
+// The skill detail page at `/skills/<name>`.
 //
-// CL-8086: this used to be a full editor over `@corbits/skills`' own
+// this used to be a full editor over `@corbits/skills`' own
 // registry — description/body editing with diff review, restore-by-
 // version off that asset's git history, a "pinned by" list, and a
 // private/shared visibility toggle. That registry (and the workflow
@@ -11,7 +11,7 @@
 // SKILL.md content, its version history, who has it pinned, or a
 // scope/visibility flag, so none of that can be rendered here without
 // vendoring a replacement surface. This page is scoped down to what the
-// stock routes actually carry until one exists — see the CL-8086 PR body
+// stock routes actually carry until one exists — see the PR body
 // for the gap.
 import { PageShell, RichEmptyState, Section, formatRelativeTime } from "@corbits/react-ui";
 import { Lightning } from "@corbits/icons";
@@ -131,12 +131,12 @@ export function SkillDetailPage({
 
       <Section
         title="Skill content"
-        description="Not readable here yet: there is no stock Interchange route for a skill's SKILL.md content (CL-8086)."
+        description="Not readable here yet: there is no stock Interchange route for a skill's SKILL.md content."
       >
         <p className="text-sm text-muted-foreground">
           This skill's description, instructions, version history, and pinned-by list lived in the
-          workbench-specific skill registry removed by CL-8086. They will return once a stock route
-          for reading and writing skill content exists.
+          workbench-specific skill registry removed by. They will return once a stock route for
+          reading and writing skill content exists.
         </p>
       </Section>
     </div>,
diff --git a/apps/web/src/pages/skills-page.tsx b/apps/web/src/pages/skills-page.tsx
index 3b467c341d..e51abc6c37 100644
--- a/apps/web/src/pages/skills-page.tsx
+++ b/apps/web/src/pages/skills-page.tsx
@@ -1,8 +1,8 @@
-// Skills: a standalone rail destination (CL-6355), over the workbench's
+// Skills: a standalone rail destination, over the workbench's
 // real skill registry (`@corbits/skills`, via `../skills-api.ts`). Used to
-// be a Settings section (CL-5990); the owner moved it back out to its own
+// be a Settings section; the owner moved it back out to its own
 // page — this is the only surface left, there is no Settings duplicate.
-// This replaced the session-local store CL-5991 shipped before that: a
+// This replaced the session-local store shipped before that: a
 // skill now lives in a native `kind:"skill"` hub asset the moment it is
 // created, and its version history is that asset's git history.
 //
@@ -54,10 +54,10 @@ function messageOf(cause: unknown): string {
 
 /**
  * The Skills roster over one workbench's skill registry, with its own
- * top-nav contract (CL-6409): the trail says where the reader is and the
+ * top-nav contract: the trail says where the reader is and the
  * top bar's action slot is the only home for "New skill". `tenantId` is the
  * registry every read is scoped to; opening a row navigates to that skill's
- * own page at `/skills/<name>` (CL-6416), which is where editing, versions,
+ * own page at `/skills/<name>`, which is where editing, versions,
  * and diffs live — this page never renders a skill inline.
  */
 export function SkillsPage({
@@ -238,7 +238,7 @@ export function SkillsPage({
 }
 
 /**
- * Skills roster mount at `/skills` (CL-6355): a thin adapter that resolves
+ * Skills roster mount at `/skills`: a thin adapter that resolves
  * which workbench's registry is listed. The stage chrome (breadcrumb trail,
  * action slot) belongs to `SkillsPage`; a single skill has its own route
  * (`/skills/<name>`, `skill-detail-page.tsx`).
diff --git a/apps/web/src/pages/workbench-timeline.tsx b/apps/web/src/pages/workbench-timeline.tsx
index 156b46d5cd..1ff15783f9 100644
--- a/apps/web/src/pages/workbench-timeline.tsx
+++ b/apps/web/src/pages/workbench-timeline.tsx
@@ -1,4 +1,4 @@
-// Per-workbench Insights view (CL-6224): one wall-clock spine merging chat
+// Per-workbench Insights view: one wall-clock spine merging chat
 // messages, thread forks, and approvals for a single workbench
 // (== workbench, per docs/GLOSSARY.md), oldest to newest with day dividers.
 // No new backend — every fetch here is an existing route this app already
diff --git a/apps/web/src/pages/workflow-detail-page.tsx b/apps/web/src/pages/workflow-detail-page.tsx
index 7eaf124107..3efc075a2d 100644
--- a/apps/web/src/pages/workflow-detail-page.tsx
+++ b/apps/web/src/pages/workflow-detail-page.tsx
@@ -1,12 +1,7 @@
-// `/workflows/<definitionId>` (CL-7371) — a workflow definition's own
-// page. CL-8160: the hub-composed detail read (name, lifecycle, source
-// commit, steps, declared-vs-approved grants, credential bindings) is
-// gone with `@corbits/workflows`'s deleted `./detail/detail-route.ts` —
-// stock's `GET /workflows/definitions` (`workflow-detail-api.ts`) exposes
-// only name, description, status (`deployed` | `stopped`), current
-// version, and timestamps. This renders exactly that; see the CL-8160 PR
-// for the upstream ask to expose more (package manifest, wire projection,
-// grant snapshot) through a stock route.
+// `/workflows/<definitionId>` — a workflow definition's own page. Stock's
+// `GET /workflows/definitions` (`workflow-detail-api.ts`) exposes only
+// name, description, status (`deployed` | `stopped`), current version,
+// and timestamps, so this renders exactly that.
 import { Badge, EmptyState, PageShell } from "@corbits/react-ui";
 import type { BadgeTone } from "@corbits/react-ui";
 import { Clock, FlowArrow } from "@corbits/icons";
diff --git a/apps/web/src/path-ids.ts b/apps/web/src/path-ids.ts
index 1dd13c1d7b..a12ecdbcb0 100644
--- a/apps/web/src/path-ids.ts
+++ b/apps/web/src/path-ids.ts
@@ -1,9 +1,9 @@
 // Path-id helpers for settings and top-level roster deep links. Shared by
 // the shell (col2 section nav) and the settings sections (stage detail) so
 // neither layer owns the other. Same pattern as workbench-path.ts. Agents
-// and Skills were Settings sections for a stretch (CL-5990) at
-// `/settings/agents/:id` and `/settings/skills/:id`; CL-6353/CL-6354/
-// CL-6355 moved both back out to their own top-level rail destinations —
+// and Skills were Settings sections for a stretch at
+// `/settings/agents/:id` and `/settings/skills/:id`; //
+// moved both back out to their own top-level rail destinations —
 // `/agents/:id` and `/skills/:id`, via `agentIdFromPath`/`skillIdFromPath`
 // below.
 
@@ -68,7 +68,7 @@ export function routineSegmentFromPath(path: string): string | null {
 /** The workflow definition id `/workflows/<id>` addresses — `null` for the
  * bare prefix or a path outside it. A workflow definition has no slug of
  * its own, so — like a routine — it is addressed by its opaque id.
- * CL-8160: since the detail page now reads stock's `workflow_definition`
+ * since the detail page now reads stock's `workflow_definition`
  * list directly, this is `workflowDefinition.id`, not `asset.id` (the
  * former Workbench-composed detail route's own address space) — the
  * export name is kept to limit the diff. */
diff --git a/apps/web/src/profile-relations.ts b/apps/web/src/profile-relations.ts
index 726eef4fad..281d05300f 100644
--- a/apps/web/src/profile-relations.ts
+++ b/apps/web/src/profile-relations.ts
@@ -1,5 +1,5 @@
 // What the profile card's Message action and shared-workbenches list need,
-// resolved against real bench data (CL-5914, CL-5919). Generalizes
+// resolved against real bench data. Generalizes
 // `myra-workbench.ts`'s ensure-style reuse — find an existing 1:1 by
 // participant instead of by a fixed title, and fall back to creating one —
 // to any profile subject rather than one hardcoded agent.
diff --git a/apps/web/src/query-client.test.ts b/apps/web/src/query-client.test.ts
index 8c371c54c4..e61b213110 100644
--- a/apps/web/src/query-client.test.ts
+++ b/apps/web/src/query-client.test.ts
@@ -55,7 +55,7 @@ describe("isAuthInvalidError", () => {
   });
 });
 
-// CL-6105: a hub restarted on an empty DB mid-session (or a cookie for a
+// a hub restarted on an empty DB mid-session (or a cookie for a
 // deleted user, or a session that simply expired) shows up as a 401 on
 // whatever query or mutation happens to run next — never a dedicated
 // "session ended" event. Every consumer of this client must be routed
diff --git a/apps/web/src/query-client.ts b/apps/web/src/query-client.ts
index bff1c221ad..d9f25a08c8 100644
--- a/apps/web/src/query-client.ts
+++ b/apps/web/src/query-client.ts
@@ -115,12 +115,12 @@ export const tenantKeys = {
    * of the literal that could drift apart. */
   workbenches: (tenantId: string, kind: WorkbenchKind) => workbenchesQueryKey(tenantId, kind),
   /** The sidebar routine-activity seam (`./shell/routine-activity.ts`).
-   * CL-8087 deleted the `feed=fires` route this read, so the query resolves
-   * with no items and issues no fetch — the key is kept (not the deleted
+   * The `feed=fires` route this read is gone, so the query resolves with
+   * no items and issues no fetch — the key is kept (not the deleted
    * `/top-level-runs` path) so both sidebar mounts still share one cache
    * entry, and so a future native fires equivalent has a key to rewire. */
   routineActivity: (tenantId: string) => ["tenant", tenantId, "routine-activity"] as const,
-  /** A workbench's own timeline reads (CL-6224): `tenantId` is the owning
+  /** A workbench's own timeline reads: `tenantId` is the owning
    * bench chat's workbench-tenancy addresses these routes at (see
    * docs/workbench-tenancy.md), `workbenchId` the workbench's own id. */
   workbenchMessages: (tenantId: string, workbenchId: string) =>
diff --git a/apps/web/src/routes.tsx b/apps/web/src/routes.tsx
index 5820ec6496..0337bd464c 100644
--- a/apps/web/src/routes.tsx
+++ b/apps/web/src/routes.tsx
@@ -7,15 +7,13 @@
 // palette even when they are off the rail. Conversation deep links
 // (`/w/:workbenchId`) stay routable; `/` is the Myra land hop (ensure +
 // open her conversation) for a bench with a workbench already, or the
-// guided first-workbench describe screen for a bench with none (CL-6104)
-// — never a Home dashboard.
-// Approvals has no page — the Activity band owns them. Agents (CL-6354)
-// and Skills (CL-6355) are their own rail destinations again — they spent
-// a stretch as Settings sections (CL-5990) and `/settings/agents[/:id]` /
+// guided first-workbench describe screen for a bench with none // — never a Home dashboard.
+// Approvals has no page — the Activity band owns them. Agents // and Skills are their own rail destinations again — they spent
+// a stretch as Settings sections and `/settings/agents[/:id]` /
 // `/settings/skills[/:id]` stay routable only as redirects back here, so
 // old links and bookmarks still land somewhere real. Library was renamed
-// Files (CL-6353) at the same time it moved off `/library`, which
-// redirects the same way. Inbox is gone too (CL-6151: tasks + approvals
+// Files at the same time it moved off `/library`, which
+// redirects the same way. Inbox is gone too (: tasks + approvals
 // don't flow into workbenches); `/inbox` stays routable only as a
 // redirect to `/`.
 
@@ -100,7 +98,7 @@ const WorkflowDetailRoute = lazy(async () => ({
   default: (await import("./pages/workflow-detail-page")).WorkflowDetailRoute,
 }));
 
-/** The signed-out screen (CL-6369) — a real route, not a conditional swap:
+/** The signed-out screen — a real route, not a conditional swap:
  * any unauthenticated request for another path bounces here with `?next=`
  * so a successful sign-in returns to where the visitor meant to go. Not
  * one of `APP_ROUTES`: like `ONBOARDING_PATH`, it renders above the shell
@@ -116,7 +114,7 @@ export const ONBOARDING_PATH = "/onboarding";
 /** Settings path — sidebar footer + settings page. */
 export const SETTINGS_PATH = "/settings";
 
-/** Mission Control — the bench's dashboard (CL-6488/CL-6489). Pinned above
+/** Mission Control — the bench's dashboard. Pinned above
  * the sidebar's footer rail as its own row (see DESIGN.md's Shell &
  * Navigation section), reachable by direct URL and the command palette
  * like everything else, but deliberately off `NAV_ROUTES`: it isn't a
@@ -124,13 +122,13 @@ export const SETTINGS_PATH = "/settings";
  * view. */
 export const MISSION_CONTROL_PATH = "/mission-control";
 
-/** The template picker (CL-6342) — every "+ New workbench" affordance
+/** The template picker — every "+ New workbench" affordance
  * (sidebar, command palette) hops here first; picking a row is what
  * actually mints the workbench. Not in `NAV_ROUTES`: it has no sidebar
  * row of its own, only the "+" control and the palette reach it. */
 export const NEW_WORKBENCH_PATH = "/new";
 
-/** Detail routes are addressed by slug (CL-6412): one route path per
+/** Detail routes are addressed by slug: one route path per
  * entity, ending in this segment. A path matches only when its last
  * segment is a real slug, so `/agents/wfd_1` still resolves to the Agents
  * roster (which owns id deep links) while `/agents/triage-bot` resolves to
@@ -139,7 +137,7 @@ const SLUG_SEGMENT = "/:slug";
 
 export const AGENT_DETAIL_PATH = `${AGENTS_PATH_PREFIX}${SLUG_SEGMENT}`;
 export const SKILL_DETAIL_PATH = `${SKILLS_PATH_PREFIX}${SLUG_SEGMENT}`;
-// Plugin detail (`/plugins/:slug`) is parked with CL-6417. CL-6817 removed
+// Plugin detail (`/plugins/:slug`) is parked with. removed
 // the "still being built" stub so gallery/palette click-throughs do not
 // promise a page that is only a placeholder.
 
@@ -210,8 +208,8 @@ export type AppRoute = {
  * and the slug-addressed detail routes (`/agents/:slug`). Other routes are
  * exact path matches. A roster prefix still matches its own nested paths,
  * so the sidebar footer row stays lit on a detail screen. Plugins is exact
- * only: until CL-6417 lands a real detail page, a slug under `/plugins` is
- * unroutable rather than a stub (CL-6817).
+ * only: until lands a real detail page, a slug under `/plugins` is
+ * unroutable rather than a stub.
  */
 export function matchesRoute(routePath: string, path: string): boolean {
   if (routePath === WORKBENCH_PATH_PREFIX) {
@@ -244,7 +242,7 @@ export function matchesRoute(routePath: string, path: string): boolean {
   return routePath === path;
 }
 
-/** Bounces old `/inbox` links and bookmarks home (CL-6151: the Inbox page
+/** Bounces old `/inbox` links and bookmarks home (: the Inbox page
  * is gone — tasks and approvals don't flow into a workbench). */
 function InboxRedirect({ navigate }: { readonly navigate: (to: string) => void }) {
   useEffect(() => {
@@ -307,7 +305,7 @@ export const APP_ROUTES: readonly AppRoute[] = [
     render: () => <RoutinesRoute />,
   },
   {
-    // A workflow definition's own page (CL-7371) — no roster of its own
+    // A workflow definition's own page — no roster of its own
     // yet, only reached by a deep link (e.g. from a routine's target).
     path: WORKFLOW_DETAIL_PATH,
     label: "Workflow",
@@ -315,7 +313,7 @@ export const APP_ROUTES: readonly AppRoute[] = [
     render: (path: string) => <WorkflowDetailRoute path={path} />,
   },
   {
-    // The renamed, remounted Library page (CL-6353) — "Library" stays out
+    // The renamed, remounted Library page — "Library" stays out
     // of user-facing copy, but the underlying artifact machinery
     // (`library-page.tsx`, `libraryArtifactIdFromPath`, …) keeps its name.
     path: "/files",
@@ -324,7 +322,7 @@ export const APP_ROUTES: readonly AppRoute[] = [
     render: (path: string) => <LibraryRoute path={path} />,
   },
   {
-    // Old `/library` links and bookmarks (CL-6353's rename) land here.
+    // Old `/library` links and bookmarks (its rename) land here.
     path: "/library",
     label: "Files",
     icon: <FolderOpen />,
@@ -349,7 +347,7 @@ export const APP_ROUTES: readonly AppRoute[] = [
     ),
   },
   {
-    // Agents spent CL-5990 through CL-6354 as a Settings section — this
+    // Agents spent through as a Settings section — this
     // entry keeps old `/settings/agents[/:id]` links routable.
     path: "/settings/agents",
     label: "Agents",
@@ -371,7 +369,7 @@ export const APP_ROUTES: readonly AppRoute[] = [
     render: (_path: string, navigate: (to: string) => void) => <SkillsRoute navigate={navigate} />,
   },
   {
-    // Skills spent CL-5990 through CL-6355 as a Settings section — this
+    // Skills spent through as a Settings section — this
     // entry keeps old `/settings/skills[/:id]` links routable.
     path: "/settings/skills",
     label: "Skills",
@@ -388,7 +386,7 @@ export const APP_ROUTES: readonly AppRoute[] = [
   },
   {
     // First-run footer rail destination. No `/plugins/:slug` until
-    // CL-6417 (CL-6817 unlinked the stub).
+    // (unlinked the stub).
     path: "/plugins",
     label: "Plugins",
     icon: <PuzzlePiece />,
diff --git a/apps/web/src/routines-api.ts b/apps/web/src/routines-api.ts
index 850989a3ad..9dcc31761e 100644
--- a/apps/web/src/routines-api.ts
+++ b/apps/web/src/routines-api.ts
@@ -2,7 +2,7 @@
 // ScheduleTrigger: list, run-now, and pause/resume. Pause/resume is the
 // same agent-directory status PUT seed uses (`stopped` / `deployed`).
 //
-// CL-8160: the hub routes `listScheduledWorkflows` and
+// the hub routes `listScheduledWorkflows` and
 // `runScheduledWorkflowNow`/`listAvailableCatalogWorkflows` used to call
 // (`@corbits/workflows`'s deleted `./schedule/scheduled-route.ts`) are
 // gone. Owner ruling: schedule state derives from stock reads client-side
@@ -13,10 +13,10 @@
 // `listAvailableCatalogWorkflows`'s connection-satisfaction read has no
 // stock equivalent either. Both resolve to an empty list rather than
 // fetch a route that no longer exists (same pattern as `tenantKeys`'
-// `routineActivity` comment in `query-client.ts` for CL-8087's deleted
+// `routineActivity` comment in `query-client.ts` for its deleted
 // `feed=fires` route): the Routines roster and Available section render
 // their existing empty states until an upstream ask lands. See the
-// CL-8160 PR for exactly what is missing.
+// PR for exactly what is missing.
 
 import { type } from "arktype";
 import { useQuery } from "@tanstack/react-query";
@@ -49,19 +49,19 @@ export const AvailableCatalogWorkflow = type({
 
 export type AvailableCatalogWorkflow = typeof AvailableCatalogWorkflow.infer;
 
-/** CL-8160: no route exists at this path any more — kept as a documented
+/** no route exists at this path any more — kept as a documented
  * dead address, not a live fetch target, for any caller that still reads
  * it for logging/keys. */
 export function scheduledWorkflowsPath(tenantId: string): string {
   return `/api/tenants/${tenantId}/workflows/scheduled`;
 }
 
-/** CL-8160: no route exists at this path any more — see the file header. */
+/** no route exists at this path any more — see the file header. */
 export function availableCatalogWorkflowsPath(tenantId: string): string {
   return `/api/tenants/${tenantId}/workflows/available`;
 }
 
-/** CL-8160: always empty — see the file header for why there is no
+/** always empty — see the file header for why there is no
  * stock-derivable replacement yet. */
 export function listAvailableCatalogWorkflows(
   _tenantId: string,
@@ -73,7 +73,7 @@ export function scheduledWorkflowRunPath(tenantId: string, definitionId: string)
   return `/api/tenants/${tenantId}/workflows/scheduled/${encodeURIComponent(definitionId)}/run`;
 }
 
-/** CL-8160: always empty — see the file header for why there is no
+/** always empty — see the file header for why there is no
  * stock-derivable replacement yet. */
 export function listScheduledWorkflows(
   _tenantId: string,
@@ -81,7 +81,7 @@ export function listScheduledWorkflows(
   return Promise.resolve([]);
 }
 
-/** CL-8160: unreachable in practice — `listScheduledWorkflows` never
+/** unreachable in practice — `listScheduledWorkflows` never
  * returns a row for `onRunNow` to be called with — kept only so
  * `useRoutineActions`' shape does not need to change too. Throws rather
  * than fetching a route that no longer exists. */
@@ -91,7 +91,7 @@ export function runScheduledWorkflowNow(
 ): Promise<{ runId: string }> {
   return Promise.reject(
     new ApiQueryError(
-      "Scheduled run-now has no stock route yet (CL-8160).",
+      "Scheduled run-now has no stock route yet.",
       undefined,
       scheduledWorkflowRunPath(_tenantId, _definitionId),
     ),
diff --git a/apps/web/src/settings-groups.ts b/apps/web/src/settings-groups.ts
index 4bafb2c86a..852143df68 100644
--- a/apps/web/src/settings-groups.ts
+++ b/apps/web/src/settings-groups.ts
@@ -1,6 +1,6 @@
 // One resolver for the settings section groups this app actually shows —
 // the package's own Account/Everyone registry, unmodified. Skills was this
-// app's one Everyone splice (CL-5990); CL-6355 moved it out to its own
+// app's one Everyone splice; moved it out to its own
 // `/skills` rail destination, so there is nothing left for this app to add
 // — both `settings-page.tsx` (stage) and `pages/settings-nav.tsx` still
 // call this rather than the package's bare `resolveSettingsSectionGroups`
diff --git a/apps/web/src/shell/app-shell.tsx b/apps/web/src/shell/app-shell.tsx
index fcf103007c..f3d46840a9 100644
--- a/apps/web/src/shell/app-shell.tsx
+++ b/apps/web/src/shell/app-shell.tsx
@@ -75,7 +75,7 @@ export function AppShell({
   const canvasFocus = useCanvasColumnFocus();
   const { selectedTenantId: tenantId } = useBench();
 
-  // A text-kind artifact's save state (CL-8189: single-user editing, no
+  // A text-kind artifact's save state (: single-user editing, no
   // co-edit presence). Resets to a fresh state the moment the open
   // artifact changes so a stale "Saved · v3" from a previous artifact can
   // never leak into a newly opened one.
diff --git a/apps/web/src/shell/bench-activity.ts b/apps/web/src/shell/bench-activity.ts
index 846fee1546..4aacfb5533 100644
--- a/apps/web/src/shell/bench-activity.ts
+++ b/apps/web/src/shell/bench-activity.ts
@@ -82,7 +82,7 @@ export function useBenchActivity(tenantId: string | null): BenchActivityQuery {
   const routinesQuery = useQuery({
     queryKey: tenantKeys.routineActivity(key),
     enabled,
-    // CL-8087: `listRoutineActivity` resolves no items without fetching
+    // `listRoutineActivity` resolves no items without fetching
     // (the `feed=fires` route is deleted and the native listing has no
     // fires equivalent), so this query never fires a request — the
     // column keeps the shared key and renders the seam's honest empty
diff --git a/apps/web/src/shell/brand-mark.tsx b/apps/web/src/shell/brand-mark.tsx
index 1a623f1f4a..4b5dfe623b 100644
--- a/apps/web/src/shell/brand-mark.tsx
+++ b/apps/web/src/shell/brand-mark.tsx
@@ -1,10 +1,7 @@
-// The sidebar's brand moment, above the create/search row (CL-6089 — the
-// switcher used to occupy the footer; the mark now stands in its place at
-// the top). A static inline SVG via react-ui's `CorbitsMark`, coloured with
-// the primary orange — simple and crisp over clever: the prior
-// `DitherBackground`-driven animation rendered as a tiny pixelated square
-// at this size (its dither grid needs real screen real estate to read),
-// so that wiring is gone rather than tuned.
+// A static inline SVG via react-ui's `CorbitsMark`, coloured with the
+// primary orange — simple and crisp over clever: an animated dither
+// background renders as a tiny pixelated square at this size, since its
+// dither grid needs real screen real estate to read.
 
 import { CorbitsMark } from "@corbits/react-ui";
 
diff --git a/apps/web/src/shell/canvas-availability.tsx b/apps/web/src/shell/canvas-availability.tsx
index 1cd4b4a02c..5c06fa7106 100644
--- a/apps/web/src/shell/canvas-availability.tsx
+++ b/apps/web/src/shell/canvas-availability.tsx
@@ -14,7 +14,7 @@ import type { CanvasColumnState } from "@corbits/shell-layout";
  * / `resolveRendererKindFromMediaType`), and the content string those
  * renderers read.
  *
- * `canEdit` (CL-5958 phase 2) opts a text-kind ("doc") artifact into
+ * `canEdit` (phase 2) opts a text-kind ("doc") artifact into
  * `ArtifactTextEditor` instead of the read-only `ArtifactRenderer` —
  * defaults to `false`/absent so every existing caller keeps rendering
  * read-only with zero behavior change. The presence `/update` route's own
@@ -28,7 +28,7 @@ export type CanvasArtifactContent = {
   readonly content: string;
   readonly unavailableReason?: string;
   readonly canEdit?: boolean;
-  /** The sandboxed preview route for an `"html"`-kind artifact (CL-5879) —
+  /** The sandboxed preview route for an `"html"`-kind artifact —
    * see `ArtifactRenderProps.previewSrc`. Absent for every other renderer
    * kind, and for an HTML blob with no Library artifact id to preview. */
   readonly previewSrc?: string;
@@ -44,7 +44,7 @@ export type RoutinePanelSubject = {
   /** Always opens the editor: a specific routine (`routineId` set) or a
    * brand-new one (`routineId` omitted or `null`) — routines-page's own
    * "New routine"/"Edit" actions, "Make this a routine", the composer's
-   * `/routine` command, and "New routine in this space" (CL-6362:
+   * `/routine` command, and "New routine in this space" (:
    * browsing/running existing routines moved to the global `/routines`
    * page, so this pane no longer has a list mode). */
   readonly routineId?: string | null;
@@ -64,7 +64,7 @@ export type RoutinePanelSubject = {
    * in which case the panel falls back to this workbench's own default
    * (Myra) workbench — never mints a new one. */
   readonly workbenchId?: string;
-  /** Seeds the target picker's initial selection (CL-7356) — the
+  /** Seeds the target picker's initial selection — the
    * conversation's own single agent participant's definition asset id,
    * when the opener could resolve exactly one. Shown visibly in
    * `DefinitionTargetPicker` and freely replaceable/clearable by the
diff --git a/apps/web/src/shell/canvas-column.tsx b/apps/web/src/shell/canvas-column.tsx
index 5b764ec648..e197c667b8 100644
--- a/apps/web/src/shell/canvas-column.tsx
+++ b/apps/web/src/shell/canvas-column.tsx
@@ -1,10 +1,10 @@
 // Column 4: the optional canvas. Collapsed, it takes no space at all — the
 // main pane gets the width back — and open, it hosts targeted auxiliary
-// content: profile cards, and (CL-5938) typed artifact renderers opened
+// content: profile cards, and typed artifact renderers opened
 // from a chat artifact chip or the Library page. Primary workbench
 // conversation lives in the main stage, not here.
 //
-// CL-8189: co-edit presence is gone. A "doc"-kind, editable artifact
+// co-edit presence is gone. A "doc"-kind, editable artifact
 // renders `@corbits/artifact-ui`'s `ArtifactTextEditor` as a plain
 // single-user controlled textarea, debounced-saved through the artifacts
 // HTTP route; every other kind stays the read-only `ArtifactRenderer`. No
@@ -137,7 +137,7 @@ function messageAction(
 }
 
 /** Insert `@handle` into whichever workbench's composer is on screen — an
- * honest "nothing to mention into" toast when none is (CL-5914: no workbench
+ * honest "nothing to mention into" toast when none is (: no workbench
  * open, or the settings surface is showing instead of a conversation). */
 function mentionAction(
   profile: ProfileSubject,
@@ -258,7 +258,7 @@ function profileActions(
     },
   };
 
-  // Pause has no backing API today (CL-5884 follow-up: no workflow-run
+  // Pause has no backing API today (follow-up: no workflow-run
   // pause endpoint exists anywhere in the hub) — omitted rather than left
   // as a no-op that pretends to do something.
   if (profile.kind === "agent") {
@@ -313,10 +313,10 @@ function toProfileCardWorkbenches(
   }));
 }
 
-/** Shared workbenches between the viewer and `profile` (CL-5919) — refetched
+/** Shared workbenches between the viewer and `profile` — refetched
  * whenever the open profile changes, dropped if a later change races past
  * an in-flight fetch. Pinned skills are intentionally never populated: no
- * agent carries any real skill-attachment data yet (tracked in CL-5991), so
+ * agent carries any real skill-attachment data yet (tracked in), so
  * showing them would be fabricated, not deferred. */
 function useSharedWorkbenches(
   tenantId: string | null,
diff --git a/apps/web/src/shell/composer-insertion.tsx b/apps/web/src/shell/composer-insertion.tsx
index ccfa17f7d4..3862d92c17 100644
--- a/apps/web/src/shell/composer-insertion.tsx
+++ b/apps/web/src/shell/composer-insertion.tsx
@@ -1,7 +1,7 @@
 // A shell-level seam between two independent surfaces: the canvas column's
 // ProfileCard (this app) and the active workbench's composer (mounted deep
 // inside `@corbits/chat-ui`'s `ChatWorkspace`, a sibling tree). The profile
-// card's Mention action (CL-5914) needs to land `@handle` in whatever
+// card's Mention action needs to land `@handle` in whatever
 // composer is on screen without either side importing the other — the same
 // "shell context exposing a callback hook" shape `canvas-availability.tsx`
 // already uses for opening a profile.
diff --git a/apps/web/src/shell/context-menu/items.test.tsx b/apps/web/src/shell/context-menu/items.test.tsx
index 990be06ec0..bfe5a3deeb 100644
--- a/apps/web/src/shell/context-menu/items.test.tsx
+++ b/apps/web/src/shell/context-menu/items.test.tsx
@@ -70,7 +70,7 @@ describe("shellContextMenuFor: workbench", () => {
     expect(itemIds(menu.entries)).toEqual(["rename", "copy-link"]);
   });
 
-  test("pin dispatches WORKBENCHES_MUTATED_EVENT so the sidebar list refetches (CL-6657)", async () => {
+  test("pin dispatches WORKBENCHES_MUTATED_EVENT so the sidebar list refetches", async () => {
     const realFetch = globalThis.fetch;
     // `patchWorkbenchSettings` validates against WorkbenchSettingsResponse —
     // a partial body rejects and the success path (event + toast) never runs.
@@ -177,7 +177,7 @@ describe("shellContextMenuFor: routine", () => {
     expect(navigate).toHaveBeenCalledWith("/routines/rt-1");
   });
 
-  // CL-8160: `runScheduledWorkflowNow` no longer calls a hub route at all
+  // `runScheduledWorkflowNow` no longer calls a hub route at all
   // (`@corbits/workflows`'s scheduled-workflow routes are deleted — see
   // `apps/web/src/routines-api.ts`'s file header) — it always rejects. The
   // former "run starts" success-path test asserted an outcome that is no
@@ -228,7 +228,7 @@ describe("shellContextMenuFor: artifact", () => {
     };
     const menu = shellContextMenuFor(target, actions());
     // Parity, not eyeballing: the context menu and the bulk action bar are
-    // driven off the exact same constant (CL-6423).
+    // driven off the exact same constant.
     expect(itemIds(menu.entries)).toEqual([...LIBRARY_BULK_OPERATION_IDS]);
     expect(findItem(menu.entries, "copy-link").label).toBe("Copy link");
   });
diff --git a/apps/web/src/shell/context-menu/items.tsx b/apps/web/src/shell/context-menu/items.tsx
index b646d6132c..600f76178e 100644
--- a/apps/web/src/shell/context-menu/items.tsx
+++ b/apps/web/src/shell/context-menu/items.tsx
@@ -209,7 +209,7 @@ function insightsRunMenu(
 }
 
 /**
- * Same operation set the Files bulk action bar offers (CL-6423): copy
+ * Same operation set the Files bulk action bar offers: copy
  * every acted-on file's canonical link. `target.ids` is either the single
  * right-clicked row, or the whole active selection when the row is part of
  * one — see `SHELL_CONTEXT_MENU_TARGETS`.
diff --git a/apps/web/src/shell/context-menu/targets.ts b/apps/web/src/shell/context-menu/targets.ts
index 2ddb668153..e34eaa63e2 100644
--- a/apps/web/src/shell/context-menu/targets.ts
+++ b/apps/web/src/shell/context-menu/targets.ts
@@ -27,7 +27,7 @@ export type ShellContextMenuTarget =
        * The ids this menu acts on: the row's own id alone, or — when the
        * right-clicked row is part of a multi-row selection — every
        * selected id, so a right-click inside an active selection offers
-       * the exact same operation set as the bulk action bar (CL-6423).
+       * the exact same operation set as the bulk action bar.
        */
       readonly ids: readonly string[];
     };
diff --git a/apps/web/src/shell/docks.tsx b/apps/web/src/shell/docks.tsx
index 9bb9101d94..11d4097fff 100644
--- a/apps/web/src/shell/docks.tsx
+++ b/apps/web/src/shell/docks.tsx
@@ -1,6 +1,6 @@
 // The initials helper the sidebar's account affordance renders from. The
 // switcher dock this file used to also export (`BenchDock`) is gone — see
-// `sidebar.tsx`'s header comment (CL-6089).
+// `sidebar.tsx`'s header comment.
 
 /**
  * Initials for the identity dock's avatar, derived locally — the app is
diff --git a/apps/web/src/shell/library-artifacts.ts b/apps/web/src/shell/library-artifacts.ts
index 56143bf75e..2385199861 100644
--- a/apps/web/src/shell/library-artifacts.ts
+++ b/apps/web/src/shell/library-artifacts.ts
@@ -91,7 +91,7 @@ export async function uploadArtifactFiles(
 }
 
 /**
- * PUT `/api/tenants/:tenantId/artifacts/:id` (CL-8189) — the artifact
+ * PUT `/api/tenants/:tenantId/artifacts/:id` — the artifact
  * editor's one save path now that co-edit presence is gone. Throws with
  * status on non-2xx so the host can render an honest failed-save state
  * instead of silently pretending the write landed.
@@ -140,7 +140,7 @@ export function artifactUploadToast(names: readonly string[]): string {
 
 /**
  * The bulk/context-menu operation set this file adopts from the shared
- * selection system (CL-6423) — deliberately just the one real, already-
+ * selection system — deliberately just the one real, already-
  * shippable operation: every other candidate (delete, move, rename,
  * download) has no backend route or store method behind it yet (see
  * `packages/artifacts-hub/src/routes.ts` and `@corbits/artifacts`'
@@ -151,7 +151,7 @@ export function artifactUploadToast(names: readonly string[]): string {
  */
 export const LIBRARY_BULK_OPERATION_IDS = ["copy-link"] as const;
 
-/** `/files/a/:id` (CL-6015) — the one canonical deep link a file has. */
+/** `/files/a/:id` — the one canonical deep link a file has. */
 export function libraryArtifactDeepLink(id: string): string {
   return `${FILES_PATH_PREFIX}/a/${encodeURIComponent(id)}`;
 }
diff --git a/apps/web/src/shell/provider-health-banner.tsx b/apps/web/src/shell/provider-health-banner.tsx
index f86f254e9e..e1d5ebd0a0 100644
--- a/apps/web/src/shell/provider-health-banner.tsx
+++ b/apps/web/src/shell/provider-health-banner.tsx
@@ -1,7 +1,7 @@
-// The shell's guided-fix banner (CL-6092): shown above the stage
+// The shell's guided-fix banner: shown above the stage
 // whenever `ProviderHealthProvider` has a classified, undismissed
 // provider-health incident *and* the person is on Settings or the
-// broken room (CL-6734). It is not a sticky toast on Skills, Files,
+// broken room. It is not a sticky toast on Skills, Files,
 // Mission Control, or Plugins — recovery lives on the surface that
 // broke. "Fix it" deep-links to Plugins' connect panel for that
 // provider (see `../pages/plugins-page.tsx`'s pending-connect
@@ -9,7 +9,7 @@
 // routes to onboarding's credential step instead, since there is no
 // provider gallery worth opening yet.
 //
-// Chrome kinds (CL-6834): `unknown` (still polling), `error` (first-load
+// Chrome kinds: `unknown` (still polling), `error` (first-load
 // poll failed — not the same silence as healthy), `healthy` (ready and
 // nothing unhealthy), and `unhealthy` (guided Fix-it banner).
 
@@ -33,7 +33,7 @@ import {
 
 const PLUGINS_PATH = "/plugins";
 
-// CL-6734: recovery on Settings and the room only — never a stalker toast.
+// recovery on Settings and the room only — never a stalker toast.
 export function isProviderHealthRecoverySurface(path: string): boolean {
   return matchesRoute(SETTINGS_PATH, path) || matchesRoute(WORKBENCH_PATH_PREFIX, path);
 }
@@ -59,7 +59,7 @@ function providerDisplayName(provider: string): string {
   return descriptor?.displayName ?? provider;
 }
 
-// Fixed, pre-written copy per classified category (CL-6092) — the one
+// Fixed, pre-written copy per classified category — the one
 // place a `ProviderHealthRecord.category` becomes a sentence. Never a
 // provider's own error text: see `provider-health.ts`'s module header for
 // why only a closed enum ever reaches this far.
@@ -133,7 +133,7 @@ export function ProviderHealthBanner({ path }: { readonly path: string }) {
 
   // A ready all-clear still mounts a zero-size marker so tests (and any
   // future chrome) can tell "actually healthy" from "unknown / not yet
-  // polled" without treating empty DOM as healthy (CL-6834).
+  // polled" without treating empty DOM as healthy.
   if (chrome.kind === "healthy" && cachedVisible === null) {
     return <div data-provider-health="healthy" hidden />;
   }
diff --git a/apps/web/src/shell/provider-health-context.test.ts b/apps/web/src/shell/provider-health-context.test.ts
index c43e64531b..2e7a9cca51 100644
--- a/apps/web/src/shell/provider-health-context.test.ts
+++ b/apps/web/src/shell/provider-health-context.test.ts
@@ -73,7 +73,7 @@ describe("deriveProviderHealthBanner", () => {
   });
 });
 
-describe("nextProviderHealthPollStatus (CL-6834)", () => {
+describe("nextProviderHealthPollStatus", () => {
   test("a successful poll always lands on ready", () => {
     expect(nextProviderHealthPollStatus("unknown", "ok")).toBe("ready");
     expect(nextProviderHealthPollStatus("error", "ok")).toBe("ready");
@@ -93,7 +93,7 @@ describe("nextProviderHealthPollStatus (CL-6834)", () => {
   });
 });
 
-describe("deriveProviderHealthChrome (CL-6834)", () => {
+describe("deriveProviderHealthChrome", () => {
   const unhealthyBanner = {
     provider: "anthropic",
     category: "credential_failure" as const,
diff --git a/apps/web/src/shell/provider-health-context.tsx b/apps/web/src/shell/provider-health-context.tsx
index 31e6322e28..f6ec1da10a 100644
--- a/apps/web/src/shell/provider-health-context.tsx
+++ b/apps/web/src/shell/provider-health-context.tsx
@@ -1,4 +1,4 @@
-// The shell's provider-health state (CL-6092): polls
+// The shell's provider-health state: polls
 // `GET .../connections/provider-health` for the selected bench and turns
 // it into the one banner-worthy provider (if any), plus the Plugins deep
 // link intent that hands a provider id from the banner's "Fix it" click
@@ -12,9 +12,9 @@
 // Dismissal is per-incident, not per-provider: dismissing an unhealthy
 // provider's banner remembers the record's own `at` timestamp, so a
 // *new* failure for the same provider (a later `at`) shows the banner
-// again — matching CL-6092's "reappears on a new failure" rule.
+// again — matching its "reappears on a new failure" rule.
 //
-// Poll status (CL-6834): empty `providers` alone is not "all healthy".
+// Poll status: empty `providers` alone is not "all healthy".
 // The chrome distinguishes `unknown` (no successful poll yet), `error`
 // (first-load / never-ready poll failure), and `ready` (a snapshot has
 // landed). A failed poll after `ready` keeps last-known providers on
@@ -50,7 +50,7 @@ export type ProviderHealthBannerState = {
   readonly zeroWorkingProviders: boolean;
 };
 
-/** Whether the shell has a usable health snapshot yet (CL-6834). */
+/** Whether the shell has a usable health snapshot yet. */
 export type ProviderHealthPollStatus = "unknown" | "ready" | "error";
 
 /**
@@ -89,7 +89,7 @@ function firstUnhealthyProvider(
 }
 
 /**
- * Pure banner-selection logic (CL-6092), split out from
+ * Pure banner-selection logic, split out from
  * `ProviderHealthProvider`'s state wiring so it is unit-testable without
  * mounting React: picks the most recently reported unhealthy provider,
  * skips it if the person already dismissed *this exact incident*
@@ -116,7 +116,7 @@ export function deriveProviderHealthBanner(
 }
 
 /**
- * Next poll-status after one fetch outcome (CL-6834). Success always
+ * Next poll-status after one fetch outcome. Success always
  * becomes `ready`. Failure before any success becomes `error` (so empty
  * providers are not read as healthy). Failure after `ready` stays
  * `ready` so last-known state remains on screen for the next retry.
@@ -133,7 +133,7 @@ export function nextProviderHealthPollStatus(
 /**
  * Maps poll status + the (possibly null) unhealthy banner into chrome
  * the shell can render. Empty providers under `unknown`/`error` are
- * never `healthy` (CL-6834).
+ * never `healthy`.
  */
 export function deriveProviderHealthChrome(
   status: ProviderHealthPollStatus,
@@ -254,13 +254,13 @@ export function useProviderHealthBanner(): ProviderHealthBannerState | null {
   return useProviderHealthContext().banner;
 }
 
-/** Poll readiness (CL-6834) — consumers that used to treat a null banner
+/** Poll readiness — consumers that used to treat a null banner
  * as "all healthy" must check this: `unknown`/`error` are not healthy. */
 export function useProviderHealthStatus(): ProviderHealthPollStatus {
   return useProviderHealthContext().status;
 }
 
-/** Chrome discriminant (CL-6834) — unknown / error / healthy / unhealthy. */
+/** Chrome discriminant — unknown / error / healthy / unhealthy. */
 export function useProviderHealthChrome(): ProviderHealthChrome {
   const { status, banner } = useProviderHealthContext();
   return deriveProviderHealthChrome(status, banner);
diff --git a/apps/web/src/shell/routine-activity.ts b/apps/web/src/shell/routine-activity.ts
index 41b73017a5..73a04bf7cb 100644
--- a/apps/web/src/shell/routine-activity.ts
+++ b/apps/web/src/shell/routine-activity.ts
@@ -1,4 +1,4 @@
-// CL-8087 deleted `@corbits/run-scope`'s `/top-level-runs?feed=fires`
+// deleted `@corbits/run-scope`'s `/top-level-runs?feed=fires`
 // route — the only listing that kept a routine's fire despite it not
 // being a top-level run — and the native `GET /workflows/runs` listing
 // has no equivalent: its top-level-only predicate drops every routine
@@ -7,7 +7,7 @@
 // name here would be exactly the slug guess the Insights feed refuses
 // to do). So `listRoutineActivity` resolves no items without fetching:
 // the shell's "Running" band and Mission Control's active-run count
-// (CL-6595) honestly report no routine activity until a native fires
+// honestly report no routine activity until a native fires
 // equivalent exists, instead of deriving routine activity from
 // top-level deployment rows that are not routine fires. The
 // `RoutineActivityItem` shape is kept so both consumers keep compiling
diff --git a/apps/web/src/shell/sidebar-rows.test.ts b/apps/web/src/shell/sidebar-rows.test.ts
index 992ae8fe45..53eb827c2d 100644
--- a/apps/web/src/shell/sidebar-rows.test.ts
+++ b/apps/web/src/shell/sidebar-rows.test.ts
@@ -123,10 +123,10 @@ describe("buildSidebarRows", () => {
     expect(rows).toEqual([{ kind: "workbench", workbench: dm }]);
   });
 
-  test("every created workbench keeps its row even when each minted its own definition (CL-6621)", () => {
-    // Creation now clones a fresh definition per workbench (CL-6452), so
+  test("every created workbench keeps its row even when each minted its own definition", () => {
+    // Creation now clones a fresh definition per workbench, so
     // sibling rows for the same agent legitimately carry distinct
-    // definitionIds. The old CL-6271 collapse keyed on exactly that and
+    // definitionIds. The old collapse keyed on exactly that and
     // hid every workbench but the newest — a person creating a second
     // workbench watched the first vanish.
     const created = ["ch_new_1", "ch_new_2", "ch_new_3"].map((id, index) =>
@@ -145,7 +145,7 @@ describe("buildSidebarRows", () => {
     expect(idsOf(rows)).toEqual(["ch_new_3", "ch_new_2", "ch_new_1"]);
   });
 
-  test("two distinct agents whose slugs humanize to the same title never collapse into one row (CL-6413)", () => {
+  test("two distinct agents whose slugs humanize to the same title never collapse into one row", () => {
     const researchAnalystHyphen = workbench({
       id: "ch_research_analyst_hyphen",
       kind: "chat",
@@ -181,7 +181,7 @@ describe("buildSidebarRows", () => {
     ]);
   });
 
-  test("same-titled DMs both stay: a title is not identity (CL-6621)", () => {
+  test("same-titled DMs both stay: a title is not identity", () => {
     const olderDm = workbench({
       id: "ch_legacy_ancestor",
       kind: "chat",
diff --git a/apps/web/src/shell/sidebar-rows.ts b/apps/web/src/shell/sidebar-rows.ts
index 0ac498f798..643eb09ad1 100644
--- a/apps/web/src/shell/sidebar-rows.ts
+++ b/apps/web/src/shell/sidebar-rows.ts
@@ -31,11 +31,11 @@ function isPinned(row: SidebarRow): boolean {
  */
 export function buildSidebarRows(items: readonly Workbench[]): readonly SidebarRow[] {
   // Every row a person can see in Postgres appears here. An earlier
-  // heuristic (CL-6271) collapsed same-agent chats onto the newest
+  // heuristic collapsed same-agent chats onto the newest
   // definitionId to hide stale cross-tenant DM siblings; once creation
-  // began cloning a fresh definition per workbench (CL-6452), that
+  // began cloning a fresh definition per workbench, that
   // heuristic could no longer tell a stale sibling from a deliberately
-  // created workbench and hid every workbench but the newest (CL-6621).
+  // created workbench and hid every workbench but the newest.
   // Hiding real workbenches reads as data loss; a duplicate stale DM is
   // merely untidy. If stale siblings resurface, fix them server-side at
   // list time, not with a client-side identity guess.
diff --git a/apps/web/src/shell/sidebar.tsx b/apps/web/src/shell/sidebar.tsx
index f3173f4ece..5ae1fca791 100644
--- a/apps/web/src/shell/sidebar.tsx
+++ b/apps/web/src/shell/sidebar.tsx
@@ -8,10 +8,10 @@
 // Always present; there is no collapse affordance and no second nav column.
 // Approvals belong in the conversation, not as a standing band here.
 //
-// Inbox is gone (CL-6151, owner decision: tasks + approvals don't flow
+// Inbox is gone (owner decision: tasks + approvals don't flow
 // into workbenches).
 //
-// No bench switcher (CL-6089): a workbench IS an agent conversation now,
+// No bench switcher: a workbench IS an agent conversation now,
 // one per account, so there is nothing to switch between in the common
 // case. A multi-bench install still resolves and routes correctly (see
 // `bench-context.tsx`) — it just has no dedicated chrome slot. The one
@@ -125,11 +125,10 @@ export function Sidebar({
 
       <SidebarPanelFooter>
         {/* Footer order: Routines, Files, Skills, Agents, Plugins, then
-            Insights only when that existing read proves real
-            items, then the account row anchors everything else (settings,
-            feedback, log out) in its pop-up menu — a single footer,
-            never two stacked rows. Routines (CL-6362) is global-only here
-            — no per-workbench routines chrome remains. */}
+            Insights only when that existing read proves real items, then
+            the account row anchors everything else (settings, feedback,
+            log out) in its pop-up menu — a single footer, never two
+            stacked rows. */}
         <button
           type="button"
           className="shell-sidebar-footer-row"
diff --git a/apps/web/src/shell/workbench-list.test.ts b/apps/web/src/shell/workbench-list.test.ts
index 079c9e3d07..5f02db6685 100644
--- a/apps/web/src/shell/workbench-list.test.ts
+++ b/apps/web/src/shell/workbench-list.test.ts
@@ -115,7 +115,7 @@ describe("orderWorkbenchRows", () => {
 });
 
 describe("filterSidebarRows", () => {
-  // CL-6662: the row shows title + preview; search must match either, or a
+  // the row shows title + preview; search must match either, or a
   // query visible in the preview (e.g. "Solvora") falsely returns no matches.
   test("matches the displayed title", () => {
     const rows = [
@@ -125,7 +125,7 @@ describe("filterSidebarRows", () => {
     expect(filterSidebarRows(rows, "launch").map((r) => r.workbench.id)).toEqual(["ch_a"]);
   });
 
-  test("matches preview text even when the title does not (CL-6662)", () => {
+  test("matches preview text even when the title does not", () => {
     const rows = [
       row({
         id: "ch_solvora",
diff --git a/apps/web/src/shell/workbench-list.tsx b/apps/web/src/shell/workbench-list.tsx
index eac55defcf..19f501734d 100644
--- a/apps/web/src/shell/workbench-list.tsx
+++ b/apps/web/src/shell/workbench-list.tsx
@@ -144,7 +144,7 @@ export function orderWorkbenchRows(workbenches: readonly Workbench[]): readonly
 /**
  * Sidebar search over what the row actually shows: the displayed title and
  * the preview snippet. Title-only matching left people staring at a visible
- * preview word (CL-6662) and a "No matches" empty state. Case-insensitive
+ * preview word and a "No matches" empty state. Case-insensitive
  * substring; empty/whitespace query keeps every row.
  */
 export function filterSidebarRows(
diff --git a/apps/web/src/skill-display-name.ts b/apps/web/src/skill-display-name.ts
index 4fefe76723..c1384aef2a 100644
--- a/apps/web/src/skill-display-name.ts
+++ b/apps/web/src/skill-display-name.ts
@@ -1,8 +1,8 @@
 // Person-facing title for a skill roster/detail surface. The registry id
 // stays the kebab slug; what a person reads is either an explicit
 // `displayTitle` (or the native skill asset's `displayName`, which the
-// stock asset routes carry as of CL-8086) or a Title Case reading of that
-// slug (CL-6747).
+// stock asset routes carry as of) or a Title Case reading of that
+// slug.
 import { humanizeSlug } from "@corbits/chat-ui/wire/display-name";
 
 export function skillDisplayName(skill: {
diff --git a/apps/web/src/skills-api.ts b/apps/web/src/skills-api.ts
index 3a7052548d..0add93afe2 100644
--- a/apps/web/src/skills-api.ts
+++ b/apps/web/src/skills-api.ts
@@ -1,6 +1,6 @@
 // The Skills settings section's one seam to the hub's skill assets.
 //
-// CL-8086: the workbench-specific skill registry this used to call
+// the workbench-specific skill registry this used to call
 // (`@corbits/skills`, mounted at `/api/tenants/:tenantId/skills`) was
 // deleted — skills are native `kind:"skill"` hub assets now, listed and
 // created through the stock asset routes (`@intx/hub-api`'s
@@ -10,7 +10,7 @@
 // version history, its scope, or who has it pinned. `description` below
 // always reads back empty and `updateSkill`/`setSkillScope`/version
 // history are gone outright rather than vendored over a route that
-// doesn't exist — see the CL-8086 PR body for this gap.
+// doesn't exist — see the PR body for this gap.
 import { type } from "arktype";
 import type { ArkErrors } from "arktype";
 
@@ -136,7 +136,7 @@ export async function loadSkill(
 /**
  * Creates a skill asset's metadata only — a bare `kind:"skill"` asset
  * row and its (empty) backing repo. There is no stock route yet to
- * populate its SKILL.md body/description in the same call (CL-8086);
+ * populate its SKILL.md body/description in the same call;
  * the caller must fill those in through whatever surface eventually
  * covers skill content.
  */
diff --git a/apps/web/src/workflow-detail-api.ts b/apps/web/src/workflow-detail-api.ts
index 1d34aca609..19e837a15f 100644
--- a/apps/web/src/workflow-detail-api.ts
+++ b/apps/web/src/workflow-detail-api.ts
@@ -1,10 +1,9 @@
-// The workflow detail page's one seam to reading a definition. CL-8160:
-// the Workbench-composed hub route (`@corbits/workflows`'s former
-// `./detail/detail-route.ts`) is deleted — there is no `GET .../:id`
-// route at all, stock or otherwise (`vendor/intx/hub-api/src/routes/
-// workflow-definitions.ts` only lists, lists versions, and rolls back).
-// This walks the stock list a page at a time and stops at the matching
-// id. Wire schema and pure display helpers live in
+// The workflow detail page's one seam to reading a definition. There is
+// no `GET .../:id` route, stock or otherwise
+// (`vendor/intx/hub-api/src/routes/workflow-definitions.ts` only lists,
+// lists versions, and rolls back), so this walks the stock list a page
+// at a time and stops at the matching id. Wire schema and pure display
+// helpers live in
 // `@corbits/workflows/client`, browser-safe like `routines-api.ts`'s own
 // definitions listing — this file is fetch composition only.
 import { type } from "arktype";
diff --git a/apps/web/test/agent-detail-page.test.tsx b/apps/web/test/agent-detail-page.test.tsx
index f901ba6e51..81a3e76ed3 100644
--- a/apps/web/test/agent-detail-page.test.tsx
+++ b/apps/web/test/agent-detail-page.test.tsx
@@ -1,4 +1,4 @@
-// The agent detail page (CL-6414) at `/agents/<slug>`: the identity card,
+// The agent detail page at `/agents/<slug>`: the identity card,
 // the system-prompt editor, the skills section, recent runs, and the
 // Duplicate/Archive/Save trio in the top bar's action slot. Mounted against
 // a stubbed hub so every edit is asserted where it matters — on the request
@@ -197,7 +197,7 @@ describe("AgentDetailPage render", () => {
     expect(markup).toContain("The registry is unreachable.");
   });
 
-  test("CL-6836: skillsError is an alert above Skills, never silent empty pins", () => {
+  test("skillsError is an alert above Skills, never silent empty pins", () => {
     const markup = renderPage({ skillsError: "500: down" });
     expect(markup).toContain('role="alert"');
     expect(markup).toContain("Could not load agent skills");
diff --git a/apps/web/test/agent-skills-picker.test.tsx b/apps/web/test/agent-skills-picker.test.tsx
index 9450657ff6..1dd1f6f0d6 100644
--- a/apps/web/test/agent-skills-picker.test.tsx
+++ b/apps/web/test/agent-skills-picker.test.tsx
@@ -1,4 +1,4 @@
-// AgentSkillsPicker (CL-5920): a pin can outlive its skill — it was
+// AgentSkillsPicker: a pin can outlive its skill — it was
 // renamed or discarded — so a stale name in `selected` must still render as
 // a removable row rather than vanish silently and leave the dialog
 // unsaveable.
diff --git a/apps/web/test/agents-api.test.ts b/apps/web/test/agents-api.test.ts
index 60b7ef28d8..1ccff88b6f 100644
--- a/apps/web/test/agents-api.test.ts
+++ b/apps/web/test/agents-api.test.ts
@@ -95,7 +95,7 @@ describe("listCatalogModels", () => {
     expect(models).toEqual([modelFixture]);
   });
 
-  test("CL-6744: drops embedding-named, hf.co, and .gguf catalog rows from the picker", async () => {
+  test("drops embedding-named, hf.co, and .gguf catalog rows from the picker", async () => {
     const embedding = {
       ...modelFixture,
       id: "mdl_embed",
@@ -220,7 +220,7 @@ describe("loadAgentDirectory", () => {
     expect(directory.skillsError).toBeUndefined();
   });
 
-  test("CL-6836: a broken skills endpoint keeps the page and surfaces skillsError, never silent empty", async () => {
+  test("a broken skills endpoint keeps the page and surfaces skillsError, never silent empty", async () => {
     stubFetch((path) => {
       if (path.includes("/workflows/definitions")) {
         return json({ data: [definitionFixture], nextCursor: null });
@@ -244,7 +244,7 @@ describe("loadAgentDirectory", () => {
     expect(directory.skillsError).toMatch(/500|down/i);
   });
 
-  // CL-8087: instances come from the native tenant-scoped
+  // instances come from the native tenant-scoped
   // `GET /workflows/runs` top-level listing — the deleted `/top-level-runs`
   // route (and its `feed=fires` variant) no longer exists.
   test("reads instances from the native top-level GET /workflows/runs listing", async () => {
diff --git a/apps/web/test/agents-page.test.tsx b/apps/web/test/agents-page.test.tsx
index d1ff4a5444..a277f98f68 100644
--- a/apps/web/test/agents-page.test.tsx
+++ b/apps/web/test/agents-page.test.tsx
@@ -1,4 +1,4 @@
-// Agents roster (CL-6354, CL-6469): a flat table of every definition a
+// Agents roster: a flat table of every definition a
 // bench owns — rows, never cards — with Status/Model/Runs·7d columns and a
 // bulk-select bar. `AgentsPage` is the presentational half (same split
 // `LibraryPage`/`LibraryRoute` use in `pages.test.tsx`); no live hub here,
@@ -21,9 +21,9 @@ import type { AgentDefinitionWithDisplayName } from "../src/agents-directory";
 import type { AgentInstance } from "../src/agents-api";
 
 // `name` is the immutable kebab identifier (the URL slug); `displayName` is
-// what `withDisplayNames` (CL-6413) derives from the definition's
+// what `withDisplayNames` derives from the definition's
 // description (or, absent one, a humanized reading of `name`). The roster
-// name slot is displayName only (CL-6748) — the slug stays the URL, not a
+// name slot is displayName only — the slug stays the URL, not a
 // second line under the name.
 const triage: AgentDefinitionWithDisplayName = {
   id: "wfd_1",
@@ -102,7 +102,7 @@ describe("runsInLast7Days", () => {
   });
 });
 
-describe("agentModelSettledContent (CL-6848)", () => {
+describe("agentModelSettledContent", () => {
   test("a fetch failure is an error, never the same label as an unset model", () => {
     expect(
       agentModelSettledContent(
@@ -139,7 +139,7 @@ describe("agentModelSettledContent (CL-6848)", () => {
     ).toEqual({ kind: "model", label: "Default" });
   });
 
-  test("an unset model stays Default when the catalog winner changes (CL-6782)", () => {
+  test("an unset model stays Default when the catalog winner changes", () => {
     const unset = {
       status: "ready" as const,
       data: { name: "triage-bot" },
@@ -157,7 +157,7 @@ describe("agentModelSettledContent (CL-6848)", () => {
     ).toEqual({ kind: "model", label: "Default" });
   });
 
-  test("a stored model is unchanged when the catalog default becomes a different winner (CL-6782)", () => {
+  test("a stored model is unchanged when the catalog default becomes a different winner", () => {
     const stored = {
       status: "ready" as const,
       data: { name: "triage-bot", model: "claude-sonnet-4" },
@@ -201,7 +201,7 @@ describe("agentModelSettledContent (CL-6848)", () => {
   });
 });
 
-describe("AgentModelCellView (CL-6848)", () => {
+describe("AgentModelCellView", () => {
   test("a capabilities fetch error is visually distinct from an unset model", () => {
     const errorMarkup = renderToStaticMarkup(
       <AgentModelCellView
@@ -249,7 +249,7 @@ describe("AgentModelCellView (CL-6848)", () => {
   });
 });
 
-describe("agentRunsSettledContent (CL-6842)", () => {
+describe("agentRunsSettledContent", () => {
   test("a runs fetch failure is not the same as an honest empty history", () => {
     expect(agentRunsSettledContent("wfd_1", [], NOW, "Couldn't load run history")).toEqual({
       kind: "error",
@@ -373,7 +373,7 @@ describe("AgentsPage", () => {
     // Live carries the live dot (react-ui's StatusDot, `live` prop) —
     // the spec's liveness marker for an actively-running agent.
     expect(markup).toContain('aria-label="Live"');
-    // Badge must not shout via CSS uppercase (same as CL-6747 skills).
+    // Badge must not shout via CSS uppercase (same as skills).
     expect(markup).toContain("normal-case");
   });
 
@@ -436,7 +436,7 @@ describe("AgentsPage", () => {
     expect(markup).toContain(">Support<");
   });
 
-  // CL-6875: the roster's quick-peek panel must hop to the agent's own page
+  // the roster's quick-peek panel must hop to the agent's own page
   // at `/agents/<slug>` — panels preview enough to decide whether to open
   // the full page (DESIGN.md), never substitute for one.
   test("selected definition's panel offers a hop to /agents/<slug>", () => {
@@ -528,7 +528,7 @@ describe("AgentsPage", () => {
     expect(markup).not.toContain("data-bulk-action");
   });
 
-  test("CL-6836: skillsError is an alert, never silent 'no skills'", () => {
+  test("skillsError is an alert, never silent 'no skills'", () => {
     const markup = renderToStaticMarkup(
       <AgentsPage
         tenantId="tnt_1"
@@ -550,7 +550,7 @@ describe("AgentsPage", () => {
     expect(markup).toContain("500: down");
   });
 
-  test("CL-6836: without skillsError, the skills failure alert is absent", () => {
+  test("without skillsError, the skills failure alert is absent", () => {
     const markup = renderToStaticMarkup(
       <AgentsPage
         tenantId="tnt_1"
@@ -569,7 +569,7 @@ describe("AgentsPage", () => {
     expect(markup).not.toContain("Could not load agent skills");
   });
 
-  test("an honest empty run history shows 0, not a load-failure marker (CL-6842)", () => {
+  test("an honest empty run history shows 0, not a load-failure marker", () => {
     const markup = renderToStaticMarkup(
       <AgentsPage
         tenantId="tnt_1"
@@ -590,7 +590,7 @@ describe("AgentsPage", () => {
     expect(markup).not.toContain("Couldn't load run history");
   });
 
-  test("a runs fetch failure is distinguishable from zero instances (CL-6842)", () => {
+  test("a runs fetch failure is distinguishable from zero instances", () => {
     const markup = renderToStaticMarkup(
       <AgentsPage
         tenantId="tnt_1"
diff --git a/apps/web/test/app-error-boundary.test.tsx b/apps/web/test/app-error-boundary.test.tsx
index 0d5991f8a5..5d386c9b3d 100644
--- a/apps/web/test/app-error-boundary.test.tsx
+++ b/apps/web/test/app-error-boundary.test.tsx
@@ -1,4 +1,4 @@
-// CL-6381: a render error anywhere in the tree used to leave the reader
+// a render error anywhere in the tree used to leave the reader
 // staring at a blank white page (no error boundary existed anywhere in
 // apps/web). This pins the boundary's red/green behaviour — a throwing
 // child renders the designed EmptyState, never a blank screen — and that
@@ -56,7 +56,7 @@ describe("AppErrorBoundary", () => {
     expect(el.textContent).not.toContain("Everything is fine");
   });
 
-  // CL-6632: a render crash that never reaches a sink is undiagnosable in
+  // a render crash that never reaches a sink is undiagnosable in
   // production — the boundary must report through `reportError`
   // (`@corbits/error-sink`) and show the refId that call returns, so a
   // person hitting this can quote it back to support.
diff --git a/apps/web/test/auth.test.tsx b/apps/web/test/auth.test.tsx
index ac7800dc9e..16c0d938ed 100644
--- a/apps/web/test/auth.test.tsx
+++ b/apps/web/test/auth.test.tsx
@@ -83,7 +83,7 @@ describe("session probe", () => {
     expect(state.kind).toBe("error");
   });
 
-  // CL-6105: a session body that parses as JSON but carries no usable
+  // a session body that parses as JSON but carries no usable
   // `user` (the shape a hub restarted against an empty DB, or a cookie
   // for a since-deleted user, can answer with) must still land on login
   // — never the "connection lost" error screen, which would strand the
@@ -98,7 +98,7 @@ describe("session probe", () => {
 /** Mirrors `main.tsx`'s `Root`'s own probe-on-mount wiring, minus
  * provisioning — the minimum needed to prove a real DOM mount, driven by
  * the real `fetchSession`, never renders the shell for an invalid session.
- * `/login` is a real route now (CL-6369): an invalid session redirects
+ * `/login` is a real route now: an invalid session redirects
  * there rather than swapping in the auth screen at whatever path was
  * requested, so this mounts a real `navigate` to follow that redirect. */
 function ProbedApp() {
@@ -216,7 +216,7 @@ describe("the gate", () => {
     const markup = renderApp({ kind: "signed-in", user });
     // The one sidebar plus the account affordance — a menu (weekly usage,
     // Settings, feedback, Log out), not a plain link straight to settings
-    // (CL-6105, grown to the reference shape in CL-6132).
+    // (grown to the reference shape in).
     expect(markup).toContain('data-testid="shell-sidebar"');
     expect(markup).toContain("shell-sidebar-account-btn");
     expect(markup).toContain('aria-label="ada · Account menu"');
diff --git a/apps/web/test/bench-activity.test.tsx b/apps/web/test/bench-activity.test.tsx
index 77c646a8a0..fa1ef7c3a6 100644
--- a/apps/web/test/bench-activity.test.tsx
+++ b/apps/web/test/bench-activity.test.tsx
@@ -87,7 +87,7 @@ describe("useBenchActivity", () => {
     container.remove();
   });
 
-  // CL-8087: the routines query resolves with no items and no fetch —
+  // the routines query resolves with no items and no fetch —
   // the `feed=fires` route is deleted and the native listing has no
   // fires equivalent — so the band renders its honest empty state.
   test("fetches workbenches, chats, and agents for the selected bench; routines resolve empty", async () => {
@@ -113,7 +113,7 @@ describe("useBenchActivity", () => {
     container.remove();
   });
 
-  test("splits workbenches by kind; routines stay empty without fetching (CL-8087)", async () => {
+  test("splits workbenches by kind; routines stay empty without fetching", async () => {
     const calls: string[] = [];
     stubTenantFetch(calls, {
       workbenches: [
@@ -177,7 +177,7 @@ describe("useBenchActivity", () => {
     container.remove();
   });
 
-  // CL-6387: a workbench minted anywhere (picker, agent launch, land-hop)
+  // a workbench minted anywhere (picker, agent launch, land-hop)
   // must show up in the sidebar the moment `createWorkbench` resolves — no
   // waiting for the next unrelated refetch. `WORKBENCHES_MUTATED_EVENT` is
   // the one signal every create path shares; this proves the listener
diff --git a/apps/web/test/canvas-column.test.tsx b/apps/web/test/canvas-column.test.tsx
index a9b1ab7fcd..61e78b2331 100644
--- a/apps/web/test/canvas-column.test.tsx
+++ b/apps/web/test/canvas-column.test.tsx
@@ -1,7 +1,7 @@
 // The canvas profile card's Message action: no silent no-op when there is
 // no bench to message against, and the panel never closes ahead of the DM
 // actually resolving — matching the toast-on-unable pattern the sibling
-// Mention action already used (CL-6019).
+// Mention action already used.
 
 import { afterEach, beforeEach, describe, expect, mock, test } from "bun:test";
 
@@ -189,7 +189,7 @@ describe("canvas profile card Message action", () => {
   });
 });
 
-describe("canvas artifact pane: single-user text editing (CL-8189)", () => {
+describe("canvas artifact pane: single-user text editing", () => {
   let container: HTMLDivElement;
   let root: Root;
 
diff --git a/apps/web/test/create-agent-panel.test.tsx b/apps/web/test/create-agent-panel.test.tsx
index 56f2302d00..c98380ba6d 100644
--- a/apps/web/test/create-agent-panel.test.tsx
+++ b/apps/web/test/create-agent-panel.test.tsx
@@ -1,16 +1,13 @@
-// DOM-mounted coverage for the shared create-agent panel (CL-6074): the
+// DOM-mounted coverage for the shared create-agent panel: the
 // happy path is name-only (a purpose is an optional, quiet secondary
 // field), Advanced stays collapsed, submitting drafts a system prompt
 // via Myra before deploying, a Suggestions card runs the exact same
 // drafting flow with its name+purpose prefilled, and a failed draft
 // fails closed — the panel surfaces the failure, keeps Advanced
 // collapsed, and the disabled primary names the blocker rather than
-// silently opening Advanced or falling back to a silent template. Settings' global
-// "Agents" directory (this form's original entry point) was cut in
-// CL-6121 — agent configuration lives per-workbench now; the chat page's
-// "+ New chat" picker moved to instant creation (CL-6081) and no longer
-// opens it — see `instant-agent-create.ts`, which reuses this panel's own
-// draft-then-create calls instead of duplicating them.
+// silently opening Advanced or falling back to a silent template.
+// `instant-agent-create.ts` reuses this panel's own draft-then-create
+// calls instead of duplicating them.
 
 import { afterEach, beforeEach, describe, expect, test } from "bun:test";
 import { readFileSync } from "node:fs";
@@ -163,7 +160,7 @@ describe("CreateAgentPanel happy path", () => {
     expect(details?.open).toBe(false);
   });
 
-  // CL-6745: create-agent dialog is an agent-definition mint, not a
+  // create-agent dialog is an agent-definition mint, not a
   // workbench mint — title and CTA must use the agent noun.
   test("dialog title and primary CTA use the agent noun, not workbench", async () => {
     await mount();
@@ -380,7 +377,7 @@ describe("CreateAgentPanel drafting failure — fails closed", () => {
 });
 
 describe("agent creation entry points", () => {
-  test("the chat page never imports this form — agent creation lives outside it entirely (CL-6138)", () => {
+  test("the chat page never imports this form — agent creation lives outside it entirely", () => {
     const chatPageSource = readFileSync(
       new URL("../src/pages/chat-page.tsx", import.meta.url),
       "utf8",
@@ -389,7 +386,7 @@ describe("agent creation entry points", () => {
     expect(chatPageSource).not.toContain("instant-agent-create");
   });
 
-  test("the command palette's new-workbench opens the template picker, not an instant mint (CL-6342)", () => {
+  test("the command palette's new-workbench opens the template picker, not an instant mint", () => {
     const commandPaletteActionsSource = readFileSync(
       new URL("../src/command-palette-actions.ts", import.meta.url),
       "utf8",
diff --git a/apps/web/test/fetch-dedupe.test.tsx b/apps/web/test/fetch-dedupe.test.tsx
index 778a366d65..ee6fcebd92 100644
--- a/apps/web/test/fetch-dedupe.test.tsx
+++ b/apps/web/test/fetch-dedupe.test.tsx
@@ -1,4 +1,4 @@
-// CL-6045: the shell used to fire the same listing 4-5x per navigation
+// the shell used to fire the same listing 4-5x per navigation
 // because independent components each fetched independently instead of
 // sharing a cache. Every mount of `useBenchActivity` (the sidebar's
 // `WorkbenchList`, and any second subscriber) shares the same TanStack
@@ -77,7 +77,7 @@ function countsByMatch(calls: readonly string[], predicate: (path: string) => bo
   return calls.filter(predicate).length;
 }
 
-describe("shell listing dedupe (CL-6045)", () => {
+describe("shell listing dedupe", () => {
   test("two WorkbenchList mounts together fetch each listing exactly once", async () => {
     const calls: string[] = [];
     stubFetch(calls);
@@ -103,7 +103,7 @@ describe("shell listing dedupe (CL-6045)", () => {
     // per (tenant, kind) no matter how many bands subscribe.
     expect(countsByMatch(calls, (p) => p.includes("kind=workbench"))).toBe(1);
     expect(countsByMatch(calls, (p) => p.includes("kind=chat"))).toBe(1);
-    // CL-8087: the routines query resolves with no items and no fetch (the
+    // the routines query resolves with no items and no fetch (the
     // `feed=fires` route is deleted and the native listing has no fires
     // equivalent), so both mounts share that without any request at all.
     expect(countsByMatch(calls, (p) => p.includes("/top-level-runs"))).toBe(0);
diff --git a/apps/web/test/home-page.test.tsx b/apps/web/test/home-page.test.tsx
index c861daa18a..8017df005e 100644
--- a/apps/web/test/home-page.test.tsx
+++ b/apps/web/test/home-page.test.tsx
@@ -7,7 +7,7 @@
 // Myra's own definition to exist, then opens her DM the same way
 // "Talk to Myra" does (`openAgentDmChat`) — never `/new`, no
 // `ensureMyraWorkbench`, no second creation path. All three entries
-// CL-6081 asks for (a direct visit to `/`, `main.tsx`'s post-login
+// asks for (a direct visit to `/`, `main.tsx`'s post-login
 // `navigate("/")`, and the onboarding wizard's post-credential hand-off)
 // resolve through this exact hop, so proving HomeRoute itself lands
 // correctly in both cases proves the direct-`/` case fully; the other
@@ -157,7 +157,7 @@ describe("HomeRoute (the `/` land hop every entry point funnels through)", () =>
 
   test("a brand-new bench with zero workbenches opens Myra's one DM, not the guided picker", async () => {
     let posted: unknown;
-    // CL-8112 T1: the deleted `/api/onboarding/provisioning-status`
+    // T1: the deleted `/api/onboarding/provisioning-status`
     // route must never be requested on this flow — record every path
     // and pin the absence below, alongside the no-error-state land.
     const requestedPaths: string[] = [];
@@ -215,7 +215,7 @@ describe("HomeRoute (the `/` land hop every entry point funnels through)", () =>
   });
 });
 
-// CL-6462: what someone sees between "connect a provider" and "talking to
+// what someone sees between "connect a provider" and "talking to
 // Myra". The old answer was a bare centred "0 of 5 ready" on an empty
 // page; these pin the replacement — one warm loader, no counts, a land
 // that happens the moment Myra herself can answer, and an honest way out
@@ -224,7 +224,7 @@ describe("the wait right after connecting a provider", () => {
   /** A bench with no workbenches yet whose setup agent (Myra) deploys
    * only after `readyAfter` definition reads — everything before that is
    * the window the person spends waiting. Readiness is the native
-   * definitions read itself (CL-8112 T1): no `/api/onboarding/*`
+   * definitions read itself (T1): no `/api/onboarding/*`
    * request anywhere in this helper. */
   function benchWhereMyraArrivesAfter(readyAfter: number) {
     let definitionReads = 0;
@@ -234,7 +234,7 @@ describe("the wait right after connecting a provider", () => {
       if (path.endsWith("/chat/workbenches") && method === "GET") {
         return json({ items: [] });
       }
-      // CL-6780: while Myra is still deploying, the credential read is
+      // while Myra is still deploying, the credential read is
       // what tells "still coming online" apart from "nothing to deploy
       // with" — an active credential means the former, so serve one: the
       // no-credential case has its own test below, and this helper must
@@ -279,7 +279,7 @@ describe("the wait right after connecting a provider", () => {
     });
   }
 
-  // CL-8112 T1: the definitions read IS the readiness check now — a
+  // T1: the definitions read IS the readiness check now — a
   // failed read must surface as an honest, retryable error, never a
   // forever "preparing" spin.
   test("shows a definitions error and retries without pretending the agent is preparing", async () => {
@@ -324,7 +324,7 @@ describe("the wait right after connecting a provider", () => {
     }
 
     const text = container?.textContent ?? "";
-    // CL-6780: zero workbenches yet — this wait is for the agent, not a
+    // zero workbenches yet — this wait is for the agent, not a
     // workbench that does not exist.
     expect(text).toContain("Preparing your agent");
     expect(text).not.toContain("Getting your workbench ready");
@@ -334,7 +334,7 @@ describe("the wait right after connecting a provider", () => {
     expect(navigated).toEqual([]);
   });
 
-  test("skip with no credential lands on an honest next step instead of stuck workbench-ready copy (CL-6780)", async () => {
+  test("skip with no credential lands on an honest next step instead of stuck workbench-ready copy", async () => {
     stubFetch((path, method) => {
       if (path === "/api/me/principals") return json(PRINCIPALS_RESPONSE);
       if (path.endsWith("/chat/workbenches") && method === "GET") {
@@ -467,7 +467,7 @@ describe('a failed memberships fetch never reads as "pick from the switcher"', (
 
 describe("the other two entries land on the same `/` hop", () => {
   test("signing in navigates home (or `next=`), not a dashboard of its own", () => {
-    // CL-6369: a plain sign-in still lands on `/` — `validatedNextPath`
+    // a plain sign-in still lands on `/` — `validatedNextPath`
     // defaults there with no `next=` param — but a sign-in redirected
     // through `/login?next=...` returns to that path instead, so the
     // literal `navigate("/")` call this test used to pin no longer
diff --git a/apps/web/test/insights-route.test.tsx b/apps/web/test/insights-route.test.tsx
index 2fb3b7dab7..f59d90c788 100644
--- a/apps/web/test/insights-route.test.tsx
+++ b/apps/web/test/insights-route.test.tsx
@@ -1,5 +1,5 @@
-// CL-6062: Insights' run feed reads from the native tenant-scoped
-// `GET /workflows/runs` top-level listing (CL-8087 repointed it off the
+// Insights' run feed reads from the native tenant-scoped
+// `GET /workflows/runs` top-level listing (repointed it off the
 // deleted `/top-level-runs` route), not the dead `/me/workflows/runs`
 // (every addressed run self-anchors at creation, so that feed's
 // `anchorRunId IS NULL` filter never matched anything). These
@@ -85,7 +85,7 @@ function stubFetch(runsBody: { data: readonly unknown[] }): RecordedCall[] {
           headers: { "content-type": "application/json" },
         }),
       );
-    // CL-8087: the native tenant-scoped top-level listing. Checked after
+    // the native tenant-scoped top-level listing. Checked after
     // the dead `/me/workflows/runs` branch above — that path also contains
     // "/workflows/runs".
     if (path.includes("/workflows/runs"))
@@ -161,7 +161,7 @@ describe("InsightsRoute run feed", () => {
     expect(calls.some((c) => c.path.includes("/me/workflows/runs"))).toBe(false);
   });
 
-  // CL-8087: the native listing carries no routine attribution at all, so
+  // the native listing carries no routine attribution at all, so
   // a row without `routineId`/`routineName` must still validate and reach
   // the surface via the definition-name fallback.
   test("a native row without routine attribution still reaches the surface", async () => {
diff --git a/apps/web/test/legacy-settings-redirects.test.tsx b/apps/web/test/legacy-settings-redirects.test.tsx
index a1d5a49e2e..3c9a157516 100644
--- a/apps/web/test/legacy-settings-redirects.test.tsx
+++ b/apps/web/test/legacy-settings-redirects.test.tsx
@@ -1,7 +1,5 @@
-// Agents and Skills moved from Settings sections back to their own rail
-// routes (CL-6354/CL-6355); Library was renamed Files off its own prefix
-// (CL-6353). Old links must still land somewhere real: these components
-// bounce to the new home, preserving any deep-linked id.
+// Old links must still land somewhere real: these components bounce to
+// the new home, preserving any deep-linked id.
 
 import { afterEach, describe, expect, test } from "bun:test";
 import { act } from "react";
diff --git a/apps/web/test/library-artifact-deep-link.test.tsx b/apps/web/test/library-artifact-deep-link.test.tsx
index 17eb973f7f..11406f5f0f 100644
--- a/apps/web/test/library-artifact-deep-link.test.tsx
+++ b/apps/web/test/library-artifact-deep-link.test.tsx
@@ -1,4 +1,4 @@
-// CL-6015: a chat artifact chip's "Open in Files" affordance navigates to
+// a chat artifact chip's "Open in Files" affordance navigates to
 // `/files/a/:id` — this is the Files side of that deep link, proving
 // the route lands on the right artifact already selected, without any
 // extra click.
diff --git a/apps/web/test/library-page-selection.test.tsx b/apps/web/test/library-page-selection.test.tsx
index 086c6d0ebe..2beb44bf57 100644
--- a/apps/web/test/library-page-selection.test.tsx
+++ b/apps/web/test/library-page-selection.test.tsx
@@ -1,4 +1,4 @@
-// CL-6423: Files adopts @corbits/react-ui's selection system. These tests
+// Files adopts @corbits/react-ui's selection system. These tests
 // drive `LibraryPage` directly (it is a pure, uncontrolled-selection
 // component outside `LibraryRoute`) so selection state, the bulk action
 // bar, and the top-nav action placement can all be asserted without a
diff --git a/apps/web/test/library-upload-confirmation.test.tsx b/apps/web/test/library-upload-confirmation.test.tsx
index d470930cce..87a0b4f70f 100644
--- a/apps/web/test/library-upload-confirmation.test.tsx
+++ b/apps/web/test/library-upload-confirmation.test.tsx
@@ -38,7 +38,7 @@ function jsonResponse(body: unknown, status = 200): Response {
 }
 
 function uploadFile(container: HTMLDivElement, file: File): void {
-  // The picker is unlabeled on purpose (CL-6750) — the visible Upload
+  // The picker is unlabeled on purpose — the visible Upload
   // button is the only named control; this helper drives the ghost input.
   const input = container.querySelector('input[type="file"]') as HTMLInputElement | null;
   if (input === null) throw new Error("no upload input");
diff --git a/apps/web/test/login-routing.test.tsx b/apps/web/test/login-routing.test.tsx
index 070295e2c0..7e0ecdb2d0 100644
--- a/apps/web/test/login-routing.test.tsx
+++ b/apps/web/test/login-routing.test.tsx
@@ -1,4 +1,4 @@
-// CL-6369: the URL bar is the source of truth for where you are.
+// the URL bar is the source of truth for where you are.
 // `TestRoot` mirrors `main.tsx`'s `Root` history wiring (same pattern as
 // `test/auth.test.tsx`'s `ProbedApp`) so these tests drive real
 // `pushState`/`popstate` traffic instead of asserting against a bare
diff --git a/apps/web/test/mission-control-page.test.tsx b/apps/web/test/mission-control-page.test.tsx
index bd825d0f35..17ce1b91e5 100644
--- a/apps/web/test/mission-control-page.test.tsx
+++ b/apps/web/test/mission-control-page.test.tsx
@@ -81,7 +81,7 @@ describe("computeInFlightRows", () => {
     expect(rows[0]?.statusLabel).toBe("Running now");
   });
 
-  // Warm-keep (CL-6681 / CL-6778): Mission Control's active-run count and
+  // Warm-keep: Mission Control's active-run count and
   // in-flight table must not keep a finished fire as "Running" forever just
   // because the delivery agent is still deployed.
   test("endedAt drops a just-finished running routine from in-flight immediately", () => {
@@ -231,7 +231,7 @@ describe("MissionControlRoute", () => {
     expect(container.textContent).not.toContain("New bench");
     expect(container.textContent).toContain("Nothing waiting on you");
     expect(container.textContent).toContain("Nothing running right now");
-    // CL-8087: the In flight band is routine-feed only and the feed is
+    // the In flight band is routine-feed only and the feed is
     // stubbed, so the empty state says so instead of implying agent runs
     // appear here.
     expect(container.textContent).toContain("Routine activity has no feed");
diff --git a/apps/web/test/pages.test.tsx b/apps/web/test/pages.test.tsx
index babbf50359..6c8d30a104 100644
--- a/apps/web/test/pages.test.tsx
+++ b/apps/web/test/pages.test.tsx
@@ -1,9 +1,5 @@
 // Screens without live backing must say so: every list renders an honest
 // empty state from real (empty) hub responses, never placeholder rows.
-// Skills coverage moved to `skills-settings-section.test.tsx` (CL-5990 —
-// it became a Settings section, not a stage-only page with an injectable
-// directory prop). Agents' Settings section was cut in CL-6121 — agent
-// configuration lives per-workbench now, not as a global tab.
 
 import { describe, expect, test } from "bun:test";
 import { renderToStaticMarkup } from "react-dom/server";
@@ -23,7 +19,7 @@ describe("empty states", () => {
     expect(markup).toContain("Upload a file, or let your agents drop their work here");
   });
 
-  // CL-6750 — empty Files must tell one story: invite to add files. A
+  // Empty Files must tell one story: invite to add files. A
   // "0 files" count beside the poster reads as a second empty announcement,
   // and a labeled ghost file input twins the visible Upload button.
   test("empty library is one invitation, not count + poster + twin uploads", () => {
diff --git a/apps/web/test/plugins-page.test.tsx b/apps/web/test/plugins-page.test.tsx
index 84a2320f82..7857168855 100644
--- a/apps/web/test/plugins-page.test.tsx
+++ b/apps/web/test/plugins-page.test.tsx
@@ -1,4 +1,4 @@
-// Screen-level proof for the Plugins gallery route (CL-6090): real fetch
+// Screen-level proof for the Plugins gallery route: real fetch
 // wiring through `listPluginsForTenant` (chain-aware credential resolve,
 // not the tenant-local list route) and `../skills-api.ts`, rendered
 // through `PluginsGallery`. Card/search/tab behavior itself is covered in
@@ -182,7 +182,7 @@ async function mount(props: { readonly navigate?: (to: string) => void } = {}) {
 }
 
 /** Stands in for `ProviderHealthBanner`'s position as a sibling of the
- * routed page under `ProviderHealthProvider` (CL-6092) — fires the same
+ * routed page under `ProviderHealthProvider` — fires the same
  * `requestPluginsConnect` the banner's "Fix it" click does, without
  * pulling in the banner widget itself. */
 function DeepLinkProbe({ provider }: { readonly provider: string }) {
@@ -253,7 +253,7 @@ describe("PluginsRoute", () => {
     expect(el.textContent).toContain("weekly-digest");
   });
 
-  test("a pending connect deep link (CL-6092) opens that provider's connect panel once the gallery loads", async () => {
+  test("a pending connect deep link opens that provider's connect panel once the gallery loads", async () => {
     stubFetch();
     const el = await mountWithDeepLink("github");
 
@@ -279,12 +279,9 @@ describe("PluginsRoute", () => {
     expect(dialog?.textContent).toContain("GitHub");
   });
 
-  // CL-6272.2: a provider connector (OpenRouter, Anthropic, ...) never
-  // gets a card on this page at all — providers live only in Shared
-  // Settings' Connections section. The OAuth-connect-return-shows-
-  // connected assertion this used to cover now lives in
-  // `@corbits/settings-ui`'s own Connections suites, the one surface
-  // that still offers a provider a Connect button to return from.
+  // A provider connector (OpenRouter, Anthropic, ...) never gets a card
+  // on this page at all — providers live only in Shared Settings'
+  // Connections section.
   test("a fresh mount never shows a provider connector, even one resolved as connected", async () => {
     globalThis.fetch = ((input: RequestInfo | URL) => {
       const path = typeof input === "string" ? input : String(input);
@@ -311,7 +308,7 @@ describe("PluginsRoute", () => {
     expect(el.textContent).not.toContain("OpenRouter");
   });
 
-  // CL-6092: a deep link naming a provider with no matching gallery card
+  // a deep link naming a provider with no matching gallery card
   // (a stale or misconfigured provider id) used to silently no-op — the
   // gallery should say something instead of nothing.
   test("a pending connect deep link with no matching gallery card renders a visible notice", async () => {
@@ -397,7 +394,7 @@ describe("PluginsRoute", () => {
     expect(navigated).toContain("/skills/summarize");
   });
 
-  // CL-7138: a fast bench switch used to let the previous tenant's
+  // a fast bench switch used to let the previous tenant's
   // in-flight fetch resolve after the new tenant's, overwriting its data.
   test("a fast bench switch never lets the previous tenant's late fetch overwrite the new one's data", async () => {
     function deferredResponse() {
@@ -523,7 +520,7 @@ describe("PluginsRoute", () => {
     expect(container.textContent).not.toContain("alpha-only");
   });
 
-  // CL-7138: cancellation must not turn every imperative reload (a
+  // cancellation must not turn every imperative reload (a
   // connect/disconnect panel's `onChanged`, the error screen's Retry) into
   // a full teardown to the loading skeleton — only a genuine tenant change
   // should do that. This drives the real disconnect flow end to end.
@@ -620,7 +617,7 @@ describe("PluginsRoute", () => {
     expect(el.textContent).not.toContain("Connected here");
   });
 
-  // CL-7141: `request_connection`'s fallback link (`/plugins?connect=<id>`)
+  // `request_connection`'s fallback link (`/plugins?connect=<id>`)
   // hands off through the same `requestPluginsConnect` path the shell
   // banner's "Fix it" click uses.
   test("a `?connect=<id>` URL naming a known connector opens that connector's connect panel", async () => {
@@ -636,7 +633,7 @@ describe("PluginsRoute", () => {
     expect(el).not.toBeNull();
   });
 
-  // CL-7141: an id the registry doesn't recognize (typo, stale link) is
+  // an id the registry doesn't recognize (typo, stale link) is
   // ignored — no dialog, no "couldn't find that connection" notice.
   test("a `?connect=<id>` URL naming an unknown connector is ignored", async () => {
     stubFetch();
@@ -649,7 +646,7 @@ describe("PluginsRoute", () => {
     expect(window.location.search).toBe("");
   });
 
-  // CL-7141: `?connect=<id>` strips only the `connect` param — any other
+  // `?connect=<id>` strips only the `connect` param — any other
   // query param this route was opened with must survive.
   test("a `?connect=<id>` URL keeps every other query param", async () => {
     stubFetch();
@@ -660,7 +657,7 @@ describe("PluginsRoute", () => {
     expect(window.location.search).toBe("?foo=bar");
   });
 
-  // CL-7141: `presetDeepLink` in `tools/connections/src/tool.ts`
+  // `presetDeepLink` in `tools/connections/src/tool.ts`
   // emits `/plugins?connect=mcp:<slug>` for a curated MCP preset (Exa,
   // Granola, Linear, ...) — this page resolves that against the preset
   // catalog and focuses the matching card's own Connect button once it
@@ -693,7 +690,7 @@ describe("PluginsRoute", () => {
     expect(window.location.search).toBe("");
   });
 
-  // CL-7141: a preset slug the catalog doesn't recognize (typo, stale
+  // a preset slug the catalog doesn't recognize (typo, stale
   // link) is ignored — no dialog, no notice, no thrown error.
   test("a `?connect=mcp:<slug>` URL naming an unknown preset is ignored", async () => {
     stubFetch();
@@ -707,7 +704,7 @@ describe("PluginsRoute", () => {
   });
 });
 
-// CL-6487: `plugins-page.tsx`'s visibility/focus refresh effect re-reads
+// `plugins-page.tsx`'s visibility/focus refresh effect re-reads
 // plugin status on `visibilitychange`/`focus` and every 30s while visible,
 // gated on a selected tenant, with a microtask guard collapsing a same-tick
 // visibilitychange+focus pair into a single reload.
diff --git a/apps/web/test/provider-health-banner.test.tsx b/apps/web/test/provider-health-banner.test.tsx
index e54068dc22..946337a707 100644
--- a/apps/web/test/provider-health-banner.test.tsx
+++ b/apps/web/test/provider-health-banner.test.tsx
@@ -1,4 +1,4 @@
-// Integration coverage for CL-6092's shell banner: mounts the real
+// Integration coverage for its shell banner: mounts the real
 // `ProviderHealthProvider` above `ProviderHealthBanner` and `PluginsRoute`
 // exactly as `app.tsx`'s `Shell` nests them (siblings under one provider,
 // same shape `shell-chrome-wiring.test.tsx` proved for canvas state),
@@ -97,7 +97,7 @@ async function flush(): Promise<void> {
 }
 
 /** Same flush loop, but waits for a specific `data-provider-health` marker
- * (CL-6834) — used when the chrome settles without an alert (healthy) or
+ * — used when the chrome settles without an alert (healthy) or
  * with an error marker rather than an unhealthy-provider alert. */
 async function flushForHealthMarker(marker: string): Promise<void> {
   for (let i = 0; i < 20; i += 1) {
@@ -116,7 +116,7 @@ function findByText(container: HTMLElement, text: string): HTMLElement | undefin
   );
 }
 
-describe("ProviderHealthBanner (CL-6092)", () => {
+describe("ProviderHealthBanner", () => {
   let container: HTMLDivElement;
   let root: Root;
 
@@ -179,7 +179,7 @@ describe("ProviderHealthBanner (CL-6092)", () => {
     expect(container.textContent).toContain("says this key is out of credit.");
   });
 
-  // CL-6092: the record only ever carries a closed category, so there is
+  // the record only ever carries a closed category, so there is
   // no provider prose left to leak through this render layer at all — but
   // this still proves it end to end, from a stubbed HTTP response through
   // to what actually lands in the DOM, in case a future category ever
@@ -259,7 +259,7 @@ describe("ProviderHealthBanner (CL-6092)", () => {
     expect(container.querySelector('[data-provider-health="healthy"]')).not.toBeNull();
   });
 
-  // CL-6834: a failed first poll used to leave providers at {}, which the
+  // a failed first poll used to leave providers at {}, which the
   // chrome treated the same as "ready and nothing unhealthy" — so an
   // unreachable health endpoint looked like every provider was fine.
   test("first-load poll failure shows error chrome, not a silent healthy state", async () => {
@@ -295,7 +295,7 @@ const CREDENTIAL_FAILURE_HEALTH = {
   connectedProviderCount: 1,
 };
 
-describe("isProviderHealthRecoverySurface (CL-6734)", () => {
+describe("isProviderHealthRecoverySurface", () => {
   test("Skills, Files, Mission Control, and Plugins are not recovery surfaces", () => {
     expect(isProviderHealthRecoverySurface(SKILLS_PATH_PREFIX)).toBe(false);
     expect(isProviderHealthRecoverySurface(`${SKILLS_PATH_PREFIX}/drafting`)).toBe(false);
@@ -314,7 +314,7 @@ describe("isProviderHealthRecoverySurface (CL-6734)", () => {
   });
 });
 
-describe("ProviderHealthBanner scope (CL-6734)", () => {
+describe("ProviderHealthBanner scope", () => {
   let container: HTMLDivElement;
   let root: Root;
 
diff --git a/apps/web/test/provisioning-error-page.test.tsx b/apps/web/test/provisioning-error-page.test.tsx
index da4fe607ef..0f4b06c3c9 100644
--- a/apps/web/test/provisioning-error-page.test.tsx
+++ b/apps/web/test/provisioning-error-page.test.tsx
@@ -1,4 +1,4 @@
-// CL-6360 regression: the onboarding failure path must render the
+// regression: the onboarding failure path must render the
 // consumer-language card — never the raw error text (freshness-gate
 // prose, file paths, stack detail) a hub route's exception happened to
 // carry. `ProvisioningErrorPage` is what `App` renders for a signed-in
diff --git a/apps/web/test/query-client.test.ts b/apps/web/test/query-client.test.ts
index 3cfe92d49d..459eb29d13 100644
--- a/apps/web/test/query-client.test.ts
+++ b/apps/web/test/query-client.test.ts
@@ -28,7 +28,7 @@ describe("pathToQueryKey", () => {
 });
 
 describe("tenantKeys.routineActivity", () => {
-  // CL-8087: the sidebar routine-activity seam keeps a shared cache key that
+  // the sidebar routine-activity seam keeps a shared cache key that
   // names the seam — not the deleted `/top-level-runs` route — so both
   // mounts subscribe to one entry and a future native fires equivalent has
   // a key to rewire.
diff --git a/apps/web/test/routes.test.tsx b/apps/web/test/routes.test.tsx
index b96db5c5a2..fc43de4d4d 100644
--- a/apps/web/test/routes.test.tsx
+++ b/apps/web/test/routes.test.tsx
@@ -22,11 +22,10 @@ import {
 } from "../src/routes";
 import type { SessionState } from "../src/session";
 
-/** Slug-addressed detail routes (CL-6412). The generic render loop below
- * renders each `route.path` verbatim, which for these is the pattern
- * (`/agents/:slug`) rather than a real path - they get their own render
- * tests instead. Plugin detail (`/plugins/:slug`) is intentionally absent
- * until CL-6417 — the stub was unlinked in CL-6817. */
+/** Slug-addressed detail routes. The generic render loop below renders
+ * each `route.path` verbatim, which for these is the pattern
+ * (`/agents/:slug`) rather than a real path — they get their own render
+ * tests instead. */
 const DETAIL_ROUTE_PATHS = new Set([
   ROUTINE_DETAIL_PATH,
   WORKFLOW_DETAIL_PATH,
@@ -34,10 +33,7 @@ const DETAIL_ROUTE_PATHS = new Set([
   SKILL_DETAIL_PATH,
 ]);
 
-/** Legacy routes that only redirect - `/library` bounces to `/files`
- * (CL-6353), `/settings/agents` and `/settings/skills` bounce to `/agents`
- * and `/skills` (CL-6354/CL-6355 moved both off Settings), `/inbox` bounces
- * home (the Inbox page is gone, CL-6151) - none has a stable panel title to
+/** Legacy routes that only redirect: none has a stable panel title to
  * assert via the generic render loop below. */
 const LEGACY_REDIRECT_PATHS = new Set([
   "/library",
@@ -204,7 +200,7 @@ describe("route table", () => {
     expect(NAV_ROUTES.map((route) => route.path)).not.toContain("/library");
   });
 
-  test("a slug segment resolves to the entity's own detail route (CL-6412)", () => {
+  test("a slug segment resolves to the entity's own detail route", () => {
     expect(matchesRoute(AGENT_DETAIL_PATH, "/agents/triage-bot")).toBe(true);
     expect(matchesRoute(SKILL_DETAIL_PATH, "/skills/pr-review")).toBe(true);
     expect(matchesRoute(ROUTINE_DETAIL_PATH, "/routines/weekly-digest")).toBe(true);
@@ -231,12 +227,12 @@ describe("route table", () => {
     expect(routeFor("/routines/rtn_1")).toBe(ROUTINE_DETAIL_PATH);
   });
 
-  test("the Plugins roster owns only its bare path until CL-6417 lands a detail page (CL-6817)", () => {
+  test("the Plugins roster owns only its bare path", () => {
     const routeFor = (path: string) =>
       APP_ROUTES.find((candidate) => matchesRoute(candidate.path, path))?.path;
     expect(routeFor("/plugins")).toBe("/plugins");
     // No stub detail: a slug under /plugins is unroutable, not a
-    // "still being built" placeholder (CL-6817).
+    // "still being built" placeholder.
     expect(routeFor("/plugins/linear")).toBeUndefined();
     expect(routeFor("/plugins/Linear")).toBeUndefined();
     expect(routeFor("/plugins/linear/settings")).toBeUndefined();
@@ -253,7 +249,7 @@ describe("route table", () => {
     expect(matchesRoute(ROUTINE_DETAIL_PATH, "/routines/%E0%A4%A")).toBe(false);
     expect(matchesRoute(AGENT_DETAIL_PATH, "/agents/%2Ftriage-bot")).toBe(false);
     // A workflow detail path reuses workflowDefinitionAssetIdFromPath
-    // (CL-7371 review): a malformed percent-escape segment must never
+    // (review): a malformed percent-escape segment must never
     // match the route at all, not match and then fail to resolve an id.
     expect(matchesRoute(WORKFLOW_DETAIL_PATH, "/workflows/%E0%A4%A")).toBe(false);
   });
@@ -297,7 +293,7 @@ describe("/inbox redirect", () => {
     container = null;
   });
 
-  test("bounces old /inbox links home - the Inbox page is gone (CL-6151)", async () => {
+  test("bounces old /inbox links home - the Inbox page is gone", async () => {
     const inboxRoute = APP_ROUTES.find((route) => route.path === "/inbox");
     if (inboxRoute === undefined) throw new Error("no /inbox route entry");
     const navigated: string[] = [];
@@ -342,7 +338,7 @@ describe("routes render", () => {
     });
   }
 
-  // Agents is the one detail route whose real screen has landed (CL-6414),
+  // Agents is the one detail route whose real screen has landed,
   // so it titles itself with the slug and lights its roster row without a
   // placeholder's "Back to" affordance.
   test("/agents/<slug> titles the agent's own page with its roster row lit", async () => {
@@ -359,9 +355,9 @@ describe("routes render", () => {
     expect(activeFooterLabel(markup)).toBe("Skills");
   });
 
-  // CL-6817: the plugin detail stub ("still being built") is gone until
-  // CL-6417 ships a real page. A slug under /plugins must not promise one.
-  test("/plugins/<slug> is not-found, never a still-being-built stub (CL-6817)", async () => {
+  // the plugin detail stub ("still being built") is gone until
+  // ships a real page. A slug under /plugins must not promise one.
+  test("/plugins/<slug> is not-found, never a still-being-built stub", async () => {
     const markup = await renderApp("/plugins/linear");
     expect(markup).toContain("Page not found");
     expect(markup).not.toContain("still being built");
diff --git a/apps/web/test/routine-activity.test.ts b/apps/web/test/routine-activity.test.ts
index 742d7694b1..42afd5ca69 100644
--- a/apps/web/test/routine-activity.test.ts
+++ b/apps/web/test/routine-activity.test.ts
@@ -1,4 +1,4 @@
-// CL-8087: `listRoutineActivity` resolves no items without fetching. The
+// `listRoutineActivity` resolves no items without fetching. The
 // `feed=fires` route it used to read is deleted with `@corbits/run-scope`,
 // and the native `GET /workflows/runs` listing has no fires equivalent —
 // its top-level-only predicate drops every routine fire by construction,
diff --git a/apps/web/test/routine-detail-page.test.tsx b/apps/web/test/routine-detail-page.test.tsx
index 69211075ab..18f14aa626 100644
--- a/apps/web/test/routine-detail-page.test.tsx
+++ b/apps/web/test/routine-detail-page.test.tsx
@@ -281,7 +281,7 @@ describe("RoutineDetailRoute", () => {
     window.localStorage.clear();
   }
 
-  // CL-8160: `listScheduledWorkflows` (`../src/routines-api.ts`) now
+  // `listScheduledWorkflows` (`../src/routines-api.ts`) now
   // always resolves empty — the hub route this mocked
   // (`/api/tenants/:id/workflows/scheduled`) is deleted with no
   // stock-derivable replacement yet. The "renders the scheduled workflow
diff --git a/apps/web/test/routines-page.test.tsx b/apps/web/test/routines-page.test.tsx
index 44e798117c..1e39e4b764 100644
--- a/apps/web/test/routines-page.test.tsx
+++ b/apps/web/test/routines-page.test.tsx
@@ -185,13 +185,13 @@ describe("GlobalRoutinesList", () => {
   });
 });
 
-// CL-8160: the "RoutinesRoute — membership-based aggregation" suite that
+// the "RoutinesRoute — membership-based aggregation" suite that
 // lived here drove `RoutinesRoute` against a mocked
 // `/api/tenants/:id/workflows/scheduled` fetch — that route
 // (`@corbits/workflows`'s deleted `./schedule/scheduled-route.ts`) is
 // gone, and `listScheduledWorkflows` (`../src/routines-api.ts`) now
 // always resolves empty rather than fetching a route that no longer
 // exists. Deleted rather than adapted: there is nothing left to mock
-// against, and no stock-derivable replacement yet (see the CL-8160 PR).
+// against, and no stock-derivable replacement yet (see the PR).
 // `GlobalRoutinesList`'s own presentational tests above are untouched —
 // only the network-backed aggregation is gone.
diff --git a/apps/web/test/settings-access.test.tsx b/apps/web/test/settings-access.test.tsx
index 30a290dec3..19f2da0938 100644
--- a/apps/web/test/settings-access.test.tsx
+++ b/apps/web/test/settings-access.test.tsx
@@ -1,6 +1,6 @@
 // The app's cached settings-access probe must agree with
 // `@corbits/settings-ui`'s mapping: evaluate effect !== allow is deny;
-// a thrown probe is error, not deny (CL-6829).
+// a thrown probe is error, not deny.
 
 import { afterEach, describe, expect, test } from "bun:test";
 import { act } from "react";
diff --git a/apps/web/test/shell-chrome-wiring.test.tsx b/apps/web/test/shell-chrome-wiring.test.tsx
index 53009b35d4..d2b84fa0a9 100644
--- a/apps/web/test/shell-chrome-wiring.test.tsx
+++ b/apps/web/test/shell-chrome-wiring.test.tsx
@@ -128,7 +128,7 @@ function Harness() {
   );
 }
 
-describe("palette actions reach the real shell state (CL-5936 sibling-context regression)", () => {
+describe("palette actions reach the real shell state (sibling-context regression)", () => {
   let container: HTMLDivElement;
   let root: Root;
 
@@ -236,7 +236,7 @@ function RouteHarness({ path }: { readonly path: string }) {
   );
 }
 
-describe("canvas dismisses when in-app nav leaves a rail surface (CL-6819)", () => {
+describe("canvas dismisses when in-app nav leaves a rail surface", () => {
   let container: HTMLDivElement;
   let root: Root;
 
diff --git a/apps/web/test/sidebar.test.tsx b/apps/web/test/sidebar.test.tsx
index a9f1e0ce44..cff98b6835 100644
--- a/apps/web/test/sidebar.test.tsx
+++ b/apps/web/test/sidebar.test.tsx
@@ -138,7 +138,7 @@ describe("Sidebar", () => {
     expect(markup).not.toContain('aria-label="Expand sidebar"');
   });
 
-  // CL-6342: the "+" control opens the template picker (`/new`) instead of
+  // the "+" control opens the template picker (`/new`) instead of
   // minting a workbench directly.
   test("the + control opens the template picker, not an instant mint", async () => {
     stubFetch();
@@ -286,7 +286,7 @@ describe("Sidebar", () => {
     container.remove();
   });
 
-  // CL-6178: the global pages (Plugins, Insights) used to be the one place
+  // the global pages (Plugins, Insights) used to be the one place
   // the sidebar dropped the workbench list — reaching a conversation from
   // there took an extra hop back through `/`. The list is not page-scoped
   // (see `workbench-list.tsx`'s own header comment), so it renders exactly
@@ -595,12 +595,12 @@ describe("Sidebar", () => {
     container.remove();
   });
 
-  // CL-6105: the footer avatar used to be a plain link straight to
+  // the footer avatar used to be a plain link straight to
   // settings — there was no way to sign out short of the hidden
   // right-click context menu. It is now a real menu (react-ui's `Menu`
   // primitives).
   //
-  // CL-6132: grown to the reference shape — the whole account row (avatar
+  // grown to the reference shape — the whole account row (avatar
   // + name) is the trigger, and the menu itself carries a weekly usage
   // line, a feedback link out to the repo's GitHub issues, a divider, and
   // a danger-styled "Log out".
@@ -657,7 +657,7 @@ describe("Sidebar", () => {
       container.remove();
     });
 
-    // CL-8160: the "Weekly usage" line read packages/insights' `/usage`
+    // the "Weekly usage" line read packages/insights' `/usage`
     // route, deleted along with the package — the menu is just feedback and
     // log out now.
     test("offers a feedback link and Log out, no Weekly usage line", async () => {
diff --git a/apps/web/test/skill-detail-page.test.tsx b/apps/web/test/skill-detail-page.test.tsx
index 635cff30fc..aab4e98e40 100644
--- a/apps/web/test/skill-detail-page.test.tsx
+++ b/apps/web/test/skill-detail-page.test.tsx
@@ -1,4 +1,4 @@
-// The skill detail page at /skills/<name> (CL-6416, rescoped by CL-8086):
+// The skill detail page at /skills/<name> (rescoped by):
 // the workbench-specific skill registry is gone, so the page renders what
 // the stock skill-asset routes carry — the display title, the updated note,
 // and an honest placeholder where the SKILL.md content, version history,
@@ -103,7 +103,7 @@ describe("SkillDetailPage", () => {
   test("renders the display title with its updated note and the content placeholder", async () => {
     stubRoutes({ [`GET ${LIST_PATH}`]: { status: 200, body: [TRIAGE_ASSET] } });
     const el = await mount({});
-    // Title is the display name, not the raw kebab slug (CL-6747).
+    // Title is the display name, not the raw kebab slug.
     expect(el.querySelector("h1")?.textContent?.trim()).toBe("Triage");
     expect(el.textContent).toContain("Updated");
     // The editor, version history, pins, and scope toggle lived in the
diff --git a/apps/web/test/skills-page.test.tsx b/apps/web/test/skills-page.test.tsx
index 68357ef494..e691ec3220 100644
--- a/apps/web/test/skills-page.test.tsx
+++ b/apps/web/test/skills-page.test.tsx
@@ -1,4 +1,4 @@
-// Skills (CL-6355), over the native skill assets (CL-8086). The page reads
+// Skills, over the native skill assets. The page reads
 // the stock asset routes (`GET /api/tenants/:id/assets?kind=skill` for the
 // roster, `POST` for create), so every case here stubs `fetch` at that seam
 // — no live hub, and no workbench-local skill registry (that path is gone).
@@ -143,7 +143,7 @@ describe("SkillsPage", () => {
       [`GET ${LIST_PATH}`]: [TRIAGE_ASSET],
     });
     const el = await mount();
-    // Name slot is a display title, never the raw kebab slug (CL-6747).
+    // Name slot is a display title, never the raw kebab slug.
     expect(el.textContent).toContain("Triage");
     // Description and scope lived in the deleted registry: the roster must
     // not invent them from asset metadata.
diff --git a/apps/web/test/stage-chrome-consistency.test.tsx b/apps/web/test/stage-chrome-consistency.test.tsx
index a170e8d0a9..0d124884f0 100644
--- a/apps/web/test/stage-chrome-consistency.test.tsx
+++ b/apps/web/test/stage-chrome-consistency.test.tsx
@@ -1,4 +1,4 @@
-// CL-6368: /files, /skills, /agents must use the same stage chrome as the
+// /files, /skills, /agents must use the same stage chrome as the
 // reference pages (Insights, Plugins) — the shared `StageTopBar` component
 // and `Table` row idiom, not bespoke divs standing in for either. This is
 // a screenshot-free assertion that each page's presentational component
@@ -32,7 +32,7 @@ afterEach(() => {
   container = null;
 });
 
-describe("stage chrome consistency (CL-6368)", () => {
+describe("stage chrome consistency", () => {
   test("Agents uses the shared StageTopBar and Table row idiom", () => {
     const markup = renderToStaticMarkup(
       <AgentsPage
diff --git a/apps/web/test/stage-search-filter.test.tsx b/apps/web/test/stage-search-filter.test.tsx
index 12a83921ec..b8de4b8da2 100644
--- a/apps/web/test/stage-search-filter.test.tsx
+++ b/apps/web/test/stage-search-filter.test.tsx
@@ -1,6 +1,6 @@
 // DECISIONS.md → Search: the stage top bar's magnifier filters the page it
 // is on — it is not a door into the global command palette, and it never
-// was meant to be one after CL-6487/CL-6410 conflated the two (PR #246).
+// was meant to be one after / conflated the two (PR #246).
 // This suite covers the per-page filter surface in isolation: a page hands
 // in its own `value`/`onChange`, and the magnifier morphs into a plain
 // input that drives that state directly, never the palette's open store.
diff --git a/apps/web/test/stage-top-nav.test.tsx b/apps/web/test/stage-top-nav.test.tsx
index 3582658a14..86369df185 100644
--- a/apps/web/test/stage-top-nav.test.tsx
+++ b/apps/web/test/stage-top-nav.test.tsx
@@ -1,4 +1,4 @@
-// CL-6409: the shell's top-nav contract. A page declares where it sits with
+// the shell's top-nav contract. A page declares where it sits with
 // `StageTopBar`'s `crumbs` — a title trail whose every level above the
 // current page is a real route — and puts its primary controls in the
 // `actions` slot. This suite covers the trail's markup and navigation, the
diff --git a/apps/web/test/toast-single-system.test.tsx b/apps/web/test/toast-single-system.test.tsx
index a2cdab8545..3d57bb0117 100644
--- a/apps/web/test/toast-single-system.test.tsx
+++ b/apps/web/test/toast-single-system.test.tsx
@@ -1,4 +1,4 @@
-// CL-6372: a failed workbench create used to fire two toasts — the house
+// a failed workbench create used to fire two toasts — the house
 // `<Toaster />` from `@corbits/react-ui` (mounted in main.tsx) and a second,
 // unstyled `<Toaster />` imported straight from `sonner` (mounted in
 // app.tsx). Sonner's `toast()` renders into every mounted `<Toaster />`, so
@@ -143,7 +143,7 @@ function emptyChannelButton(): HTMLButtonElement | null {
   return container?.querySelector(".new-workbench-empty-channel") ?? null;
 }
 
-describe("the one toast system (CL-6372)", () => {
+describe("the one toast system", () => {
   // The store outlives this file too: a sibling suite that raised a toast
   // before bun loaded this one leaves it queued, and it would render into
   // the first `<Toaster />` mounted here. Start every test from an empty
@@ -173,7 +173,7 @@ describe("the one toast system (CL-6372)", () => {
     await waitForClear();
   });
 
-  // CL-6510: the new contract this file's own change introduced — a
+  // the new contract this file's own change introduced — a
   // missing setup agent no longer fires a toast at all, since the
   // picker now shows a retryable "still setting up" panel instead of
   // treating that precondition as a dead end.
@@ -186,7 +186,7 @@ describe("the one toast system (CL-6372)", () => {
       if (path.includes("/workflows/definitions")) {
         return Promise.resolve(json({ data: [], nextCursor: null }));
       }
-      // CL-8112 T1: no `/api/onboarding/*` request anywhere on this
+      // T1: no `/api/onboarding/*` request anywhere on this
       // path — the empty native definition alone drives the retry panel.
       return Promise.resolve(json({ error: "boom" }, 500));
     }) as typeof fetch;
diff --git a/apps/web/test/workbench-list.test.tsx b/apps/web/test/workbench-list.test.tsx
index 96b24a30b1..c006a7756c 100644
--- a/apps/web/test/workbench-list.test.tsx
+++ b/apps/web/test/workbench-list.test.tsx
@@ -145,7 +145,7 @@ describe("WorkbenchList — pending-approvals signal", () => {
   });
 });
 
-describe("WorkbenchList — pin visibility and order (CL-6657)", () => {
+describe("WorkbenchList — pin visibility and order", () => {
   test("pin glyph floats across the mixed list, never within Agents/Channels sections", async () => {
     stubFetch({
       chats: [
diff --git a/apps/web/test/workflow-detail-page.test.tsx b/apps/web/test/workflow-detail-page.test.tsx
index a76a2c5889..bf598fc8d5 100644
--- a/apps/web/test/workflow-detail-page.test.tsx
+++ b/apps/web/test/workflow-detail-page.test.tsx
@@ -1,4 +1,4 @@
-// `/workflows/<definitionId>` (CL-7371, thinned CL-8160): a workflow
+// `/workflows/<definitionId>` (thinned): a workflow
 // definition's own page. Covers the pure `WorkflowDetailPage` body
 // against fixtures for what stock's `GET /workflows/definitions` (via
 // `getWorkflowDefinitionDetail`) actually exposes — name, description,
diff --git a/scripts/db-setup.ts b/scripts/db-setup.ts
index 1fdb9285d6..b62a4009b6 100644
--- a/scripts/db-setup.ts
+++ b/scripts/db-setup.ts
@@ -52,7 +52,7 @@ const INSTALLED_PACKAGE_MIGRATIONS: readonly {
   { name: "@corbits/webhook-triggers", apply: applyWebhookTriggersMigrations },
   { name: "@corbits/notify", apply: applyNotifyMigrations },
   { name: "@corbits/mailbox", apply: applyMailboxMigrations },
-  // Own `inbox` schema; CL-8185 forward-drops the CL-7208 snooze table.
+  // Own `inbox` schema; forward-drops the snooze table.
   { name: "@corbits/inbox", apply: applyInboxMigrations },
   { name: "@corbits/agent-directory", apply: applyAgentDirectoryMigrations },
   { name: "@corbits/cron", apply: applyCronMigrations },
@@ -82,7 +82,7 @@ async function applyInstalledPackageMigrations(databaseUrl: string): Promise<voi
     }
   }
 
-  // CL-7593: the one-time CL-7242 duplicate-grant cleanup ran its era
+  // the one-time duplicate-grant cleanup ran its era
   // (the repo_review_lease now prevents new duplicates at write time),
   // so db-setup no longer performs it: setup applies migrations only,
   // and grant state stays on the native grant routes.
@@ -367,11 +367,11 @@ async function dropRoutinesSchemaAfterDigestHandoff(databaseUrl: string): Promis
 
 /**
  * One-shot, forward-only: `@corbits/run-key-history` and
- * `@corbits/preferences` (CL-8158) had zero web callers on their mounts —
+ * `@corbits/preferences` had zero web callers on their mounts —
  * a diagnostics-only listener and a client that was built but never
  * imported — so both packages, their mounts, and their schemas are gone.
- * `bench` (CL-8160) joins them for the same reason. `insights` (also
- * CL-8160) joins them too, for a different reason: the hub mounts nothing
+ * `bench` joins them for the same reason. `insights` (also
+ *) joins them too, for a different reason: the hub mounts nothing
  * Workbench-specific, so its usage/latency writers and read routes are
  * gone — Insights UI now reads stock observability/workflow routes
  * client-side instead. Absent schema is a no-op; safe to re-run.
diff --git a/scripts/e2e/browser/walkthrough.ts b/scripts/e2e/browser/walkthrough.ts
index faa5701ff6..34b5a80b3a 100644
--- a/scripts/e2e/browser/walkthrough.ts
+++ b/scripts/e2e/browser/walkthrough.ts
@@ -3,7 +3,7 @@
 // bodies only — they run inside the browser tab, not this Bun process, but
 // TypeScript still checks them against this file's own lib config.
 //
-// Real-browser acceptance harness (CL-6072): boots the real stack (a fresh
+// Real-browser acceptance harness: boots the real stack (a fresh
 // scratch database, the real hub, the real sidecar, a stubbed provider
 // credential) and drives the actual built web app in headless Chrome via
 // puppeteer-core — clicking, typing, and screenshotting exactly as a human
@@ -391,7 +391,7 @@ async function closeCreateAgentDrawer(page: Page): Promise<void> {
 
 /**
  * Drives the one creation flow a person actually walks: the sidebar's "+"
- * opens `/new` (prompt-primary picker, CL-8156 — a plain tenant + Myra,
+ * opens `/new` (prompt-primary picker, — a plain tenant + Myra,
  * no template cards). Clicking the empty-channel button mints
  * immediately. Waits for the URL to land on a fresh `/w/:id` distinct
  * from wherever the click started.
@@ -563,7 +563,7 @@ async function run(): Promise<void> {
       () => page,
       "02-onboarding-provisioning",
       async () => {
-        // CL-6089 dropped the naming step: provisioning now fires
+        // dropped the naming step: provisioning now fires
         // automatically, under a default name derived from the account, the
         // moment the wizard mounts. This step just waits for it to land on
         // the credential step next — an inference-provider radiogroup.
@@ -581,7 +581,7 @@ async function run(): Promise<void> {
       () => page,
       "03-connect-provider-stub-key",
       async () => {
-        // The hub never seeds (CL-8207): the onboarding page's own
+        // The hub never seeds: the onboarding page's own
         // `ProviderConnectStep` is the only path that plants a credential
         // and installs Myra, so this step drives that real form — the
         // same radiogroup + password input a person fills in — rather
@@ -612,7 +612,7 @@ async function run(): Promise<void> {
       },
     );
 
-    // --- Step 2: CL-7053 — `/` opens Myra's DM (kind chat, titled Myra).
+    // --- Step 2: — `/` opens Myra's DM (kind chat, titled Myra).
     // A brand-new account's bare root lands in that DM — not an auto-minted
     // empty workbench, not `/new`, no describe screen. Plus/picker still
     // mints empty "New Workbench" channels (step 05). Keep waiting for `/w/`
@@ -704,7 +704,7 @@ async function run(): Promise<void> {
         // only replaces the empty state once `/api/me/principals` resolves
         // and picks a tenant. Poll rather than reading once, so this never
         // flakes on that ordinary reload race. The rail is one mixed list
-        // of agent DMs and channels (CL-7053) — do not require labeled
+        // of agent DMs and channels — do not require labeled
         // "Agents" / "Channels" headings. Presence of Myra's row is the
         // mixed-list signal.
         let myraRows = 0;
@@ -736,7 +736,7 @@ async function run(): Promise<void> {
       },
     );
 
-    // --- Step 3: CL-6342 — the sidebar's "+" opens `/new`. Clicking
+    // --- Step 3: — the sidebar's "+" opens `/new`. Clicking
     // "Just start talking" mints an empty channel immediately (no kind
     // radio, no Create step). The mint must land somewhere NEW, distinct
     // from Myra's DM.
@@ -770,9 +770,9 @@ async function run(): Promise<void> {
         // The always-visible sidebar already lists every conversation —
         // the rows are just there, no navigation or priming needed. First
         // land is Myra's DM (titled Myra). The picker mint (05) is the one
-        // empty "New Workbench" channel — plus does not clone the DM
-        // (CL-7053). The list is a cache invalidated by the create event,
-        // so give the refetch a bounded moment to land before judging.
+        // empty "New Workbench" channel — plus does not clone the DM. The
+        // list is a cache invalidated by the create event, so give the
+        // refetch a bounded moment to land before judging.
         await page.waitForSelector(".shell-ch-row-wrap", { timeout: 15_000 });
         let mintedRows = 0;
         let myraRows = 0;
@@ -859,7 +859,7 @@ async function run(): Promise<void> {
         }
         if (live) {
           // Myra hosts every workbench and speaks first: her greeting must
-          // land without the person typing anything (CL-6126/CL-6137).
+          // land without the person typing anything.
           const greeted = await page
             .waitForSelector('div.chat-bubble-row[data-own="false"]', {
               timeout: 60_000,
@@ -971,7 +971,7 @@ async function run(): Promise<void> {
       },
     );
 
-    // --- Step 6: CL-6067 / CL-6069 — hub restart, stale-thread reconnect
+    // --- Step 6: / — hub restart, stale-thread reconnect
     await step(
       () => page,
       "09-restart-hub-same-db",
@@ -997,7 +997,7 @@ async function run(): Promise<void> {
 
     await step(
       () => page,
-      "10-reload-open-existing-chat-CL-6067-6069",
+      "10-reload-open-existing-chat--6069",
       async () => {
         await page.goto(`${webBaseUrl}/c`, {
           waitUntil: "domcontentloaded",
@@ -1016,7 +1016,7 @@ async function run(): Promise<void> {
             status: "repro-confirmed",
             detail:
               "after restart, the reloaded shell never regained the picker-minted New Workbench row within 15s " +
-              "(stuck disconnected) — CL-6067 reproduced",
+              "(stuck disconnected) — reproduced",
           };
         }
         await clickStable(
@@ -1033,7 +1033,7 @@ async function run(): Promise<void> {
           return {
             status: "repro-confirmed",
             detail:
-              'opening the existing chat after restart shows "Couldn\'t load" — CL-6069 ' +
+              'opening the existing chat after restart shows "Couldn\'t load" — ' +
               "stale-thread failure reproduced",
           };
         }
@@ -1054,9 +1054,9 @@ async function run(): Promise<void> {
       },
     );
 
-    // --- Step 11 (CL-7366): the routines page's target picker must offer
-    // at least one real option (CL-7351 target discovery) and creation
-    // must stay blocked until one is picked (CL-7355) — the panel has no
+    // --- Step 11: the routines page's target picker must offer
+    // at least one real option (target discovery) and creation
+    // must stay blocked until one is picked — the panel has no
     // Save button, it autosaves on blur once a target is chosen.
     await step(
       () => page,
diff --git a/scripts/e2e/db-gate.test.ts b/scripts/e2e/db-gate.test.ts
index ef217f360e..555cafe006 100644
--- a/scripts/e2e/db-gate.test.ts
+++ b/scripts/e2e/db-gate.test.ts
@@ -2,7 +2,7 @@
 // total silence. Locally (CI unset) it returns describe.skip and the
 // summary names `compose.test.yml`. Under CI=true it throws
 // instead, so a miswired pipeline cannot report green on a suite that
-// never ran. None of ci.yml's jobs provision Postgres (CL-8150); they
+// never ran. None of ci.yml's jobs provision Postgres; they
 // all still skip with the same compose hint.
 
 import { afterEach, describe, expect, test } from "bun:test";
diff --git a/scripts/e2e/db-gate.ts b/scripts/e2e/db-gate.ts
index 7b82b57c9c..117a58397f 100644
--- a/scripts/e2e/db-gate.ts
+++ b/scripts/e2e/db-gate.ts
@@ -17,7 +17,7 @@ export const MISSING_DATABASE_HINT =
   `DATABASE_URL=${TEST_DATABASE_URL} bun test ...`;
 
 // GitHub Actions sets CI=true on every job. None of ci.yml's jobs
-// provision Postgres (CL-8150 dropped the db-backed e2e/isolation/
+// provision Postgres (dropped the db-backed e2e/isolation/
 // db-suites jobs), so every DB-gated suite skips there (loudly). Any
 // other CI context — including `CI=true bun test` locally — treats a
 // missing DATABASE_URL as a hard failure so a miswired pipeline cannot
diff --git a/scripts/e2e/docker-sidecar-smoke.ts b/scripts/e2e/docker-sidecar-smoke.ts
index 93f26863b7..51a2be4204 100644
--- a/scripts/e2e/docker-sidecar-smoke.ts
+++ b/scripts/e2e/docker-sidecar-smoke.ts
@@ -1,4 +1,4 @@
-// CL-6096: drives the real @corbits/docker-provisioner plugin against a
+// drives the real @corbits/docker-provisioner plugin against a
 // live Docker daemon — create -> exclusive-run (container appears) ->
 // release -> container gone. This is the one proof of the docker sidecar
 // provisioner that needs real infrastructure, so it lives outside `bun
diff --git a/scripts/e2e/harness.test.ts b/scripts/e2e/harness.test.ts
index 0664a0faa6..64165e019d 100644
--- a/scripts/e2e/harness.test.ts
+++ b/scripts/e2e/harness.test.ts
@@ -1,5 +1,5 @@
 // Coverage for the cleanup registry every e2e/smoke suite in this
-// directory shares (CL-5515): `tempDir` really mkdtemps under the OS
+// directory shares: `tempDir` really mkdtemps under the OS
 // temp dir with the given prefix, and `track` accepts a spawned app
 // without requiring its `stop()` to fire immediately (that only
 // happens once the suite's own tests finish, via `afterAll`).
@@ -126,7 +126,7 @@ describe("parseEnvFileDatabaseUrl", () => {
   });
 });
 
-// CL-7472 regression: at pin 692c3106, the hub authenticates a
+// regression: at pin 692c3106, the hub authenticates a
 // sidecar's register/reconnect frame only against a `sidecar` row that
 // resolves to a live `sidecar_allocation` (or `workflow_probe`) row —
 // upstream dropped the older "shared" credential scope this repo used
diff --git a/scripts/e2e/harness.ts b/scripts/e2e/harness.ts
index 4a9763f285..41c1425ce2 100644
--- a/scripts/e2e/harness.ts
+++ b/scripts/e2e/harness.ts
@@ -143,7 +143,7 @@ export async function connectE2eDb(databaseUrl: string): Promise<SqlClient> {
  * a sidecar identity that resolves to a live `sidecar_allocation` (or
  * `workflow_probe`) row — the standalone, unallocated ("shared"
  * credential scope) sidecar this used to provision no longer
- * authenticates at all (CL-7472), so a caller of this function gets a
+ * authenticates at all, so a caller of this function gets a
  * connect-and-immediately-reject loop, not a usable dial-in. Every
  * suite this repo runs against CI now lets the hub's own
  * process-provisioner spawn a dedicated sidecar per allocation instead
@@ -273,7 +273,7 @@ export async function startHub(options: {
     // The routine scheduler's real production cadence is a 30s
     // setInterval (routine-scheduler.ts); waiting that out for real on
     // every e2e test that touches a scheduled routine fire is exactly
-    // the sleep-dominated cost CL-7250 exists to cut. The default and
+    // the sleep-dominated cost exists to cut. The default and
     // the override are both proven in-process with fake timers in
     // apps/hub/test/routine-scheduler.test.ts, so no e2e test needs the
     // real cadence — a caller's extraEnv can still opt back into it by
@@ -317,7 +317,7 @@ export async function startHub(options: {
  * Boot a standalone sidecar as a real process, pointed at the hub's
  * WebSocket dial-in route with a `provisionSidecar`-issued identity.
  * See `provisionSidecar`'s own doc comment: at pin 692c3106 this
- * identity never authenticates (CL-7472), so the process this starts
+ * identity never authenticates, so the process this starts
  * connects and is immediately rejected in a loop rather than dialing
  * in. Kept only for the scripts not yet migrated to the hub's
  * automatic, per-allocation process-provisioner spawn.
@@ -410,7 +410,7 @@ export function expectStatus(what: string, result: ApiResult, expected: number):
  * each e2e suite chooses per file whether it stays zero-network: a
  * suite that pins every inference source at its own local noop
  * inference server (`./noop-inference-server.ts`; the hub itself mounts
- * no such route as of CL-8160) or an unreachable placeholder host
+ * no such route as of) or an unreachable placeholder host
  * imports this guard and calls it at every baseURL/apiKey it
  * constructs, so an accidental live-provider reference fails
  * immediately instead of silently attempting a real call. A suite that
diff --git a/scripts/e2e/noop-inference-server.ts b/scripts/e2e/noop-inference-server.ts
index 84574a57e2..2096d05813 100644
--- a/scripts/e2e/noop-inference-server.ts
+++ b/scripts/e2e/noop-inference-server.ts
@@ -1,10 +1,8 @@
-// A standalone stand-in for a real Anthropic endpoint, moved here from
-// the hub (`apps/hub/src/e2e/noop-inference.ts`, CL-8160): a scheduled
-// step that needs no model must not call one, so the hub no longer
-// mounts this route at all. An e2e suite that wants a deployment costing
-// zero real inference now starts this tiny server itself
-// (`startNoopInferenceServer`) and plants a catalog offering pointed at
-// its own `baseUrl` (`ensureNoopCatalogOffering`), never at the hub.
+// A standalone stand-in for a real Anthropic endpoint: an e2e suite that
+// wants a deployment costing zero real inference starts this tiny server
+// itself (`startNoopInferenceServer`) and plants a catalog offering
+// pointed at its own `baseUrl` (`ensureNoopCatalogOffering`), never at
+// the hub.
 //
 // The wire shape matches exactly what
 // `@intx/inference/src/providers/anthropic.ts`'s `parseResponse` accepts
@@ -127,7 +125,7 @@ const NOOP_OFFERING_PRIORITY = 0;
  * Plants (or finds) the dedicated catalog offering an e2e suite deploys
  * a zero-cost workflow against: a catalog model named `NOOP_MODEL`, a
  * provider pointed at this file's own `startNoopInferenceServer`
- * `baseUrl` (never the hub — CL-8160 dropped the hub's own mount), and a
+ * `baseUrl` (never the hub — dropped the hub's own mount), and a
  * placeholder credential — the same `ensureCatalogModel` / `ensureProvider`
  * / `ensureCredential` / `ensureCatalogProvider` / `ensureCatalogOffering`
  * sequence `seedCatalog` runs for a real provider, run here for this one
diff --git a/scripts/e2e/smoke-auth.test.ts b/scripts/e2e/smoke-auth.test.ts
index 09ba0fde15..012ef5a046 100644
--- a/scripts/e2e/smoke-auth.test.ts
+++ b/scripts/e2e/smoke-auth.test.ts
@@ -1,4 +1,4 @@
-// Smoke scenario 1/5 (CL-6004): sign-up + session. The narrowest proof
+// Smoke scenario 1/5: sign-up + session. The narrowest proof
 // in the suite — a real hub process, a real better-auth email sign-up,
 // and the resulting session cookie actually authorizing a session-gated
 // route. No tenant, no sidecar: this only proves the auth boundary
diff --git a/scripts/e2e/smoke-chat.test.ts b/scripts/e2e/smoke-chat.test.ts
index fd1e19cc9f..f9f4fb17c6 100644
--- a/scripts/e2e/smoke-chat.test.ts
+++ b/scripts/e2e/smoke-chat.test.ts
@@ -1,4 +1,4 @@
-// Smoke scenario 3/5 (CL-6004): chat round-trip. A workbench is created,
+// Smoke scenario 3/5: chat round-trip. A workbench is created,
 // a message is posted and read back with its content intact, and the
 // workbench's invited-agent listing answers with the documented shape.
 // This deliberately stops short of `chat.test.ts`'s full battery
diff --git a/scripts/e2e/smoke-onboarding.test.ts b/scripts/e2e/smoke-onboarding.test.ts
index 6e025d2d09..cb15b786b1 100644
--- a/scripts/e2e/smoke-onboarding.test.ts
+++ b/scripts/e2e/smoke-onboarding.test.ts
@@ -1,4 +1,4 @@
-// Smoke scenario 2/5 (CL-6004): first-tenant provisioning. Stock
+// Smoke scenario 2/5: first-tenant provisioning. Stock
 // Interchange cutover: composition mounts no provisioning hook
 // (`POST /api/onboarding/provision` is gone) — signup mints nothing,
 // and the first bench comes from an ordinary `POST /api/tenants` with
diff --git a/scripts/e2e/smoke-webhook.test.ts b/scripts/e2e/smoke-webhook.test.ts
index f167234d5d..4fd49c5b29 100644
--- a/scripts/e2e/smoke-webhook.test.ts
+++ b/scripts/e2e/smoke-webhook.test.ts
@@ -1,4 +1,4 @@
-// Smoke scenario 5/5 (CL-6004): webhook trigger. A workflow is bound to
+// Smoke scenario 5/5: webhook trigger. A workflow is bound to
 // a `@corbits/webhook-triggers` row instead of a cadence; a correctly
 // HMAC-signed payload delivered to the public ingress route launches a
 // run. Zero-cost: the deployed workflow's only inference source is this
@@ -113,7 +113,7 @@ describe.skipIf(databaseUrl === undefined)("smoke: webhook trigger", () => {
 
       // The zero-cost catalog chain: an anthropic-plugin provider whose
       // base URL is this suite's own noop inference server, never the
-      // hub (CL-8160 dropped its noop-inference mount) and never a real
+      // hub (dropped its noop-inference mount) and never a real
       // model.
       const noopServer = startNoopInferenceServer();
       track({
@@ -186,7 +186,7 @@ describe.skipIf(databaseUrl === undefined)("smoke: webhook trigger", () => {
           if (hub.exited()) {
             throw new Error(`hub exited before deploy; output:\n${hub.output()}`);
           }
-          // CL-7071 cutover: the launched run's sidecar initialization
+          // cutover: the launched run's sidecar initialization
           // resolves its `@corbits` scope pins against the tenant's own
           // `corbits-tools` package registry — without it the allocation
           // fails with `sidecar_initialization_failed` and the run never
@@ -360,72 +360,69 @@ describe.skipIf(databaseUrl === undefined)("smoke: webhook trigger", () => {
         },
       );
 
-      await hop(
-        "the run's insights trace reads real spans once the run settles (CL-5910)",
-        async () => {
-          // The webhook-fired run turns inference against the noop source
-          // asynchronously (see the sidecar dial-in above). The run itself
-          // (`workflow_run`) is a long-lived folded run that stays
-          // "running" indefinitely, the same way a chat workbench does —
-          // it is the per-turn `inference_turn` row (written by
-          // @intx/hub-sessions' event-collector) that records the turn.
-          // Poll straight out of the database (no route reads a
-          // non-anchored run's turns by bare id, per the file header)
-          // until one exists — the reader must surface it regardless of
-          // whether the turn has settled yet, rendering a still-running
-          // turn as `awaiting` rather than waiting it out — then assert
-          // the mounted RunTraceReader (apps/hub/src/index.ts) surfaces a
-          // non-empty, honestly-timed trace through the real HTTP route —
-          // never the "run_trace_reader_not_mounted" absent-shape this
-          // route used to return unconditionally.
-          const sql = await connectE2eDb(url);
-          const deadline = Date.now() + 60_000;
-          try {
-            for (;;) {
-              const rows = await sql.unsafe(
-                `SELECT "status" FROM "inference_turn" WHERE "instance_id" = $1`,
-                [instanceId],
-              );
-              if (rows.length > 0) break;
-              if (Date.now() > deadline) {
-                throw new Error(`run ${instanceId} recorded no inference_turn within the deadline`);
-              }
-              await Bun.sleep(200);
+      await hop("the run's insights trace reads real spans once the run settles ", async () => {
+        // The webhook-fired run turns inference against the noop source
+        // asynchronously (see the sidecar dial-in above). The run itself
+        // (`workflow_run`) is a long-lived folded run that stays
+        // "running" indefinitely, the same way a chat workbench does —
+        // it is the per-turn `inference_turn` row (written by
+        // @intx/hub-sessions' event-collector) that records the turn.
+        // Poll straight out of the database (no route reads a
+        // non-anchored run's turns by bare id, per the file header)
+        // until one exists — the reader must surface it regardless of
+        // whether the turn has settled yet, rendering a still-running
+        // turn as `awaiting` rather than waiting it out — then assert
+        // the mounted RunTraceReader (apps/hub/src/index.ts) surfaces a
+        // non-empty, honestly-timed trace through the real HTTP route —
+        // never the "run_trace_reader_not_mounted" absent-shape this
+        // route used to return unconditionally.
+        const sql = await connectE2eDb(url);
+        const deadline = Date.now() + 60_000;
+        try {
+          for (;;) {
+            const rows = await sql.unsafe(
+              `SELECT "status" FROM "inference_turn" WHERE "instance_id" = $1`,
+              [instanceId],
+            );
+            if (rows.length > 0) break;
+            if (Date.now() > deadline) {
+              throw new Error(`run ${instanceId} recorded no inference_turn within the deadline`);
             }
-          } finally {
-            await sql.end();
+            await Bun.sleep(200);
           }
+        } finally {
+          await sql.end();
+        }
 
-          const traceRes = await api(
-            hub.baseUrl,
-            "GET",
-            `/api/tenants/${tenantId}/insights/runs/${instanceId}/trace`,
-            undefined,
-            cookies,
+        const traceRes = await api(
+          hub.baseUrl,
+          "GET",
+          `/api/tenants/${tenantId}/insights/runs/${instanceId}/trace`,
+          undefined,
+          cookies,
+        );
+        expectStatus("get run trace", traceRes, 200);
+        const trace = traceRes.data as {
+          runId: string;
+          spans: { id: string; kind: string; phase: string }[];
+        };
+        expect(trace.runId).toBe(instanceId);
+        if (trace.spans.length === 0) {
+          throw new Error(
+            `run trace came back empty for a run with a recorded turn: ${JSON.stringify(trace)}`,
           );
-          expectStatus("get run trace", traceRes, 200);
-          const trace = traceRes.data as {
-            runId: string;
-            spans: { id: string; kind: string; phase: string }[];
-          };
-          expect(trace.runId).toBe(instanceId);
-          if (trace.spans.length === 0) {
-            throw new Error(
-              `run trace came back empty for a run with a recorded turn: ${JSON.stringify(trace)}`,
-            );
-          }
-          const turnSpan = trace.spans.find((span) => span.kind === "turn");
-          if (turnSpan === undefined) {
-            throw new Error(`run trace had no "turn" span: ${JSON.stringify(trace)}`);
-          }
-          expect(["ok", "awaiting", "failed"]).toContain(turnSpan.phase);
-        },
-      );
+        }
+        const turnSpan = trace.spans.find((span) => span.kind === "turn");
+        if (turnSpan === undefined) {
+          throw new Error(`run trace had no "turn" span: ${JSON.stringify(trace)}`);
+        }
+        expect(["ok", "awaiting", "failed"]).toContain(turnSpan.phase);
+      });
 
       console.log(
         "smoke-webhook: a signed delivery to the public ingress route " +
           "launched a real run row for a webhook trigger, " +
-          "and its insights trace reads back real spans (CL-5910).",
+          "and its insights trace reads back real spans.",
       );
     }
   }, 180_000);
diff --git a/scripts/e2e/t3-github-roundtrip.proof.test.ts b/scripts/e2e/t3-github-roundtrip.proof.test.ts
index 245f25004e..8bc47b2080 100644
--- a/scripts/e2e/t3-github-roundtrip.proof.test.ts
+++ b/scripts/e2e/t3-github-roundtrip.proof.test.ts
@@ -1,4 +1,4 @@
-// CL-8114 evidence kept as a regression proof: exercises the browser's exact
+// evidence kept as a regression proof: exercises the browser's exact
 // GitHub connect→disconnect round-trip (`connections/github/complete` →
 // credentials list → `connections/github/disconnect` → credentials list)
 // against a booted hub on a scratch database with the operator's real `gh`
diff --git a/scripts/e2e/walking-skeleton.test.ts b/scripts/e2e/walking-skeleton.test.ts
index b6f51ef79d..c7635b72b0 100644
--- a/scripts/e2e/walking-skeleton.test.ts
+++ b/scripts/e2e/walking-skeleton.test.ts
@@ -66,7 +66,7 @@ function stringField(data: unknown, field: string, what: string): string {
  * model) pairs a step's own agent declares
  * (`@intx/workflow-deploy`'s `pickStepInferenceSource`); naming any other
  * model here 409s "no approved inference source" at deploy even though
- * this suite never calls real inference (CL-7473).
+ * this suite never calls real inference.
  */
 const PLACEHOLDER_MODEL = "noop";
 
@@ -287,7 +287,7 @@ describe.skipIf(databaseUrl === undefined)("walking skeleton", () => {
         if (hub.exited()) {
           throw new Error(`hub exited before deploy; output:\n${hub.output()}`);
         }
-        // CL-7071 cutover: the launched run's sidecar initialization
+        // cutover: the launched run's sidecar initialization
         // resolves its `@corbits` scope pins against the tenant's own
         // `corbits-tools` package registry — Myra's assistant definition
         // carries several such pins, so this registry must exist before
diff --git a/scripts/lib/publishable-packages.ts b/scripts/lib/publishable-packages.ts
index 711762292e..a9a4a44062 100644
--- a/scripts/lib/publishable-packages.ts
+++ b/scripts/lib/publishable-packages.ts
@@ -1,4 +1,4 @@
-// Shared plumbing for deploy-compatibility auditing (CL-8167): finds every
+// Shared plumbing for deploy-compatibility auditing: finds every
 // workspace package and splits it into publishable (not `private`) vs.
 // still-private, the same way scripts/checks/publishable-workflows.ts
 // scopes its audit. Computed from package.json files on disk — never
diff --git a/scripts/repro/connect-github-smoke.ts b/scripts/repro/connect-github-smoke.ts
index 28ffdf2a2d..f9dc535b62 100644
--- a/scripts/repro/connect-github-smoke.ts
+++ b/scripts/repro/connect-github-smoke.ts
@@ -1,6 +1,6 @@
 /**
- * Manual smoke test for the GitHub connect binding (CL-6344, hub-zero T3
- * CL-8114): connects a real PAT through the native tenant-scoped
+ * Manual smoke test for the GitHub connect binding (hub-zero T3
+ *): connects a real PAT through the native tenant-scoped
  * `connections/github/complete` route against a running local hub.
  *
  * Hub-zero T3 deleted the workbench-scoped `github/state` and
@@ -62,10 +62,10 @@ async function main() {
   if (!connectRes.ok) {
     throw new Error("connect failed — see status above");
   }
-  // Hub-zero T3 (CL-8114) stops here: `github/state` and
-  // `github/start-reviewing` no longer exist, so there is nothing further
-  // to smoke until the connections follow-up lands the native rebind.
-  console.log("connected — state/start-reviewing walkthrough deleted (CL-8114)");
+  // `github/state` and `github/start-reviewing` no longer exist, so
+  // there is nothing further to smoke until the connections follow-up
+  // lands the native rebind.
+  console.log("connected — state/start-reviewing walkthrough deleted");
 }
 
 main().catch((err: unknown) => {
diff --git a/skills/skills/src/skill-md.ts b/skills/skills/src/skill-md.ts
index f6cd066a2d..04a18d0445 100644
--- a/skills/skills/src/skill-md.ts
+++ b/skills/skills/src/skill-md.ts
@@ -113,9 +113,9 @@ export function buildSkillMd(input: {
  *
  * No backfill for the dropped `skill_access` table: every pre-cutover
  * SKILL.md lacks `scope`, and no seed, fixture, template, or
- * provisioning path plants a tenant-visible row (verified CL-7583), so
+ * provisioning path plants a tenant-visible row (verified), so
  * there is nothing that would silently narrow. Pre-GA cutovers are not
- * migrated — dev databases reseed via `bun run reset` (CL-7445 ruling;
+ * migrated — dev databases reseed via `bun run reset` (ruling;
  * see docs/package-migrations.md). */
 export function parseSkillMd(text_: string): ParsedSkillMd {
   const lines = text_.split(/\r?\n/);
diff --git a/test/isolation/isolation.test.ts b/test/isolation/isolation.test.ts
index 782343afdb..07b0542305 100644
--- a/test/isolation/isolation.test.ts
+++ b/test/isolation/isolation.test.ts
@@ -69,11 +69,8 @@ if (!databaseUrl) {
   // harness.ts`) gives its process up to 5s to exit on SIGTERM before
   // escalating to SIGKILL — so a graceful-but-slow exit on either
   // process alone can approach Bun's 5000ms default hook timeout, and
-  // two of them in sequence can exceed it outright. CL-7462: this hook
-  // timed out at ~5001ms in CI once boot wiring grew (workflow
-  // allocation/dispatch initialization, extra reconciliation queries),
-  // pushing shutdown latency on CI's shared runners past the default
-  // margin even though nothing was actually hung.
+  // two of them in sequence can exceed it outright, even though nothing
+  // is actually hung.
   afterAll(async () => {
     await hub.shutdown();
   }, 20_000);
diff --git a/tools/access/src/client.ts b/tools/access/src/client.ts
index ea0b43f293..8e1a019722 100644
--- a/tools/access/src/client.ts
+++ b/tools/access/src/client.ts
@@ -265,7 +265,7 @@ export class DelegationCeilingError extends Error {}
  * within it.
  *
  * Two fidelity gaps against a server-side check, both recorded as upstream
- * asks on CL-7575: role-derived grants are not visible through the stock
+ * asks on: role-derived grants are not visible through the stock
  * principal-filtered listing, and the read-then-write is not atomic.
  */
 export async function firstActionOutsideCeiling(
diff --git a/tools/agent-directory/src/client.test.ts b/tools/agent-directory/src/client.test.ts
index f59dea6171..e85fabc101 100644
--- a/tools/agent-directory/src/client.test.ts
+++ b/tools/agent-directory/src/client.test.ts
@@ -35,8 +35,7 @@ test("createAgentDefinition posts to the workflow-agent-directory definitions en
         name: "Research Buddy",
         description: null,
         // The real create route serializes a `text` DB column, always
-        // a string on the wire — never the JS number literal CL-6480
-        // let this schema wrongly accept.
+        // a string on the wire — never the JS number literal         // let this schema wrongly accept.
         currentVersion: "1",
         status: "deployed",
         skills: [],
diff --git a/tools/agent-directory/src/client.ts b/tools/agent-directory/src/client.ts
index d5b054e4d9..117d5d3483 100644
--- a/tools/agent-directory/src/client.ts
+++ b/tools/agent-directory/src/client.ts
@@ -42,7 +42,7 @@ export interface CreatedAgentDefinition {
   readonly description: string | null;
   /** The create route serializes `workflow_definition.current_version`,
    * a `text` DB column, verbatim — always a string on the wire, never
-   * a number (CL-6480: parsing this as `"number"` made every genuine
+   * a number (: parsing this as `"number"` made every genuine
    * success fail this schema and read as a create failure). */
   readonly currentVersion: string;
   readonly status: string;
@@ -50,7 +50,7 @@ export interface CreatedAgentDefinition {
   /** Set when `model` was requested but the tenant's catalog didn't
    * offer it, so the route substituted its default (or left the
    * definition modelless) instead of baking in a name that can never
-   * resolve (CL-6477). `null` when the requested model — or its
+   * resolve. `null` when the requested model — or its
    * absence — needed no substitution. */
   readonly modelNote: string | null;
 }
diff --git a/tools/agent-directory/src/tool.ts b/tools/agent-directory/src/tool.ts
index 88e6122f21..b5ba774a50 100644
--- a/tools/agent-directory/src/tool.ts
+++ b/tools/agent-directory/src/tool.ts
@@ -153,7 +153,7 @@ async function runCreateAgent(env: WorkflowAgentDirectoryEnv, call: ToolCall): P
   // Set when the requested `modelPreference` fell outside the tenant's
   // catalog and the route substituted its default (or left the
   // definition modelless) instead of baking in a name that can never
-  // resolve (CL-6477) — surfaced on every branch below so the model
+  // resolve — surfaced on every branch below so the model
   // relays the substitution to the user rather than claiming the
   // model it originally asked for.
   const modelSuffix = created.modelNote !== null ? ` ${created.modelNote}` : "";
diff --git a/tools/capability/README.md b/tools/capability/README.md
index 621bad4330..2bc607e101 100644
--- a/tools/capability/README.md
+++ b/tools/capability/README.md
@@ -1,6 +1,6 @@
 # @corbits/capability-tools
 
-The `request_capability` `@intx/agent` tool bundle (CL-6084): an agent's
+The `request_capability` `@intx/agent` tool bundle: an agent's
 in-chat way to ask for a tool package, skill, or model it doesn't have
 yet. Declared `approval: "ask"` — Interchange's native per-invocation
 gate suspends the call as a pending approval and renders it in chat
@@ -23,7 +23,7 @@ next reply."); an out-of-inventory request comes back naming what's
 actually available, never a fabricated success. Any other transport or
 HTTP failure comes back as an honest `isError: true` result.
 
-## Activated end-to-end (CL-6086)
+## Activated end-to-end
 
 Both gaps this package was originally written against are now closed:
 
@@ -44,8 +44,8 @@ Both gaps this package was originally written against are now closed:
    That route deliberately does NOT gate on a grant-store check for the
    own-definition case — Interchange's grant materialization still
    never seeds a `kind: "workflow"` run's principal an `update` grant on
-   its own definition ([Intx gap], tracked durably by CL-6085). Instead
-   the route relies on the approval that already happened: this
+   its own definition. Instead the route relies on the approval that
+   already happened: this
    bundle's `request_capability` tool is declared `approval: "ask"`, so
    the reactor suspends every call as a pending approval and renders it
    in-chat BEFORE this bundle's `run` — and therefore before the route
diff --git a/tools/capability/package.json b/tools/capability/package.json
index e28c42f1a8..775f304106 100644
--- a/tools/capability/package.json
+++ b/tools/capability/package.json
@@ -1,7 +1,7 @@
 {
   "name": "@corbits/capability-tools",
   "version": "0.0.7",
-  "description": "The request_capability tool as an @intx/agent tool bundle (CL-6084): an agent asks in-chat for a tool package, skill, or model it doesn't have, gated behind Interchange's native per-invocation approval, and — once approved — calls the sanctioned workflow-run capabilities surface with the run's own bearer token, never a model-supplied identity",
+  "description": "The request_capability tool as an @intx/agent tool bundle: an agent asks in-chat for a tool package, skill, or model it doesn't have, gated behind Interchange's native per-invocation approval, and — once approved — calls the sanctioned workflow-run capabilities surface with the run's own bearer token, never a model-supplied identity",
   "license": "LGPL-2.1-or-later",
   "repository": {
     "type": "git",
diff --git a/tools/capability/src/client.ts b/tools/capability/src/client.ts
index 024648e34d..adac43a24c 100644
--- a/tools/capability/src/client.ts
+++ b/tools/capability/src/client.ts
@@ -1,26 +1,15 @@
 // A minimal client for the workflow-run-authenticated capabilities
 // surface a running agent calls to add itself a tool package, skill, or
-// model — the execution half of `@corbits/agent-directory`'s already-
-// shipped `POST /:definitionId/capabilities` (fail-closed, versioned,
-// refresh-on-success) and `GET /capabilities/inventory`.
+// model — authenticated via `createWorkflowRunAuthenticator` (sidecar
+// bearer token + run address), never a human browser session.
 //
-// This surface is `@corbits/agent-directory`'s `createWorkflowCapabilityRoutes`
-// (`packages/agent-directory/src/workflow-capability-routes.ts`, CL-6086),
-// mounted in `apps/hub` at `/api/workflow-capabilities` beside
-// `/api/workflow-skills` — authenticated the same way, via
-// `createWorkflowRunAuthenticator` (sidecar bearer token + run address),
-// never a human browser session.
-//
-// [Intx gap, tracked durably by CL-6085]: a run's own `kind: "workflow"`
-// principal is still never seeded a `workflow-definition: <its own id>/
-// update` grant anywhere in `vendor/intx/hub-api`'s grant materialization.
-// The workflow-capability route does not block on this: it skips a
-// grant-store check for the narrow own-definition case, relying instead
-// on `request_capability`'s `approval: "ask"` gate already having put a
-// human in front of the call before this client is ever invoked — see
-// the route's own file-level comment for the full authorization
-// reasoning. `requireGrant` will replace that interim rule once CL-6085
-// closes.
+// Intx gap: a run's own `kind: "workflow"` principal is never seeded a
+// `workflow-definition: <its own id>/update` grant, so the
+// workflow-capability route skips a grant-store check for the narrow
+// own-definition case, relying instead on `request_capability`'s
+// `approval: "ask"` gate having already put a human in front of the
+// call before this client is ever invoked — see the route's own
+// file-level comment for the full authorization reasoning.
 import { type } from "arktype";
 import {
   runBearerHeaders,
diff --git a/tools/capability/src/tool.ts b/tools/capability/src/tool.ts
index 29292eace4..a909fb18bd 100644
--- a/tools/capability/src/tool.ts
+++ b/tools/capability/src/tool.ts
@@ -10,7 +10,7 @@
 //
 // `definitionId` is threaded onto `env` by the sidecar's per-step env
 // builder (`apps/sidecar/src/workflow-substrate-factory/step-env.ts`,
-// CL-6086), the same ground `@corbits/memory-tools`' `hubMemoryUrl`/
+//), the same ground `@corbits/memory-tools`' `hubMemoryUrl`/
 // `sidecarToken`/`address` are threaded from — resolved at the
 // substrate factory from the deploying definition's own id
 // (`WORKFLOW_DEFINITION_REPO_ID`). `WorkflowCapabilityEnv` below
@@ -19,7 +19,7 @@
 //
 // See `./client.ts` for the workflow-run-authenticated capabilities
 // route (`@corbits/agent-directory`'s `createWorkflowCapabilityRoutes`,
-// also CL-6086) this bundle's execution calls.
+// also) this bundle's execution calls.
 import { defineTool } from "@intx/agent";
 import type { BaseEnv } from "@intx/agent";
 import type { ToolCall, ToolResult } from "@intx/types/runtime";
@@ -144,7 +144,7 @@ async function runRequestCapability(
 
 /**
  * The `@corbits/capability-tools` bundle factory: one tool,
- * `approval: "ask"`, four env keys — the CL-6084 self-service capability
+ * `approval: "ask"`, four env keys — the self-service capability
  * request path. `description` is written to read as an approval card
  * headline (`@corbits/approvals`' `headlineFor` uses a tool's
  * `description` verbatim, appending `arguments.title` in quotes when the
diff --git a/tools/catalog/README.md b/tools/catalog/README.md
index 83f42d42a9..841af15ae9 100644
--- a/tools/catalog/README.md
+++ b/tools/catalog/README.md
@@ -23,7 +23,7 @@ tools read only, so none is gated behind approval.
 
 ## Managing offerings
 
-Catalog administration is UI-only (CL-7588): models, providers, and
+Catalog administration is UI-only: models, providers, and
 offerings are managed through the tenant-admin catalog routes
 (`vendor/intx/hub-api/src/routes/{models,model-providers,
 model-offerings}.ts`) behind a browser session. This package exposes no
diff --git a/tools/catalog/src/index.ts b/tools/catalog/src/index.ts
index 31efc74fa3..a9fbae9776 100644
--- a/tools/catalog/src/index.ts
+++ b/tools/catalog/src/index.ts
@@ -12,5 +12,5 @@ export {
 // two stock Interchange tenant reads — `/api/tenants/:tenantId/models` and
 // `/api/tenants/:tenantId` — with the run's own bearer credential. There is
 // no `/api/workflow-inference-catalog` mount any more. Catalog
-// administration stays UI-only (CL-7588): this package exposes no write
+// administration stays UI-only: this package exposes no write
 // tools and no hub routes.
diff --git a/tools/catalog/src/tool.test.ts b/tools/catalog/src/tool.test.ts
index 7ab7f34985..1feda6cc41 100644
--- a/tools/catalog/src/tool.test.ts
+++ b/tools/catalog/src/tool.test.ts
@@ -192,5 +192,5 @@ describe("list_model_concepts", () => {
     expect(content).toContain("nothing here can do it yet");
   });
 });
-// Catalog administration is UI-only (CL-7588): no write tools, so no
+// Catalog administration is UI-only: no write tools, so no
 // write-tool tests — the read-only describes above are the whole suite.
diff --git a/tools/catalog/src/tool.ts b/tools/catalog/src/tool.ts
index dabcec1353..9944295fce 100644
--- a/tools/catalog/src/tool.ts
+++ b/tools/catalog/src/tool.ts
@@ -320,5 +320,5 @@ export const catalogTools = defineTool<WorkflowCatalogEnv>({
     },
   }),
 });
-// Catalog administration is UI-only (CL-7588): this package exposes no
+// Catalog administration is UI-only: this package exposes no
 // write tools — the three read-only tools above are the whole bundle.
diff --git a/tools/connections/src/tool.test.ts b/tools/connections/src/tool.test.ts
index 7cf3033b61..9656a09773 100644
--- a/tools/connections/src/tool.test.ts
+++ b/tools/connections/src/tool.test.ts
@@ -158,7 +158,7 @@ test("request_connection tells the agent to keep helping for a name this workspa
     expect(result.content).toMatch(/can't connect/i);
     expect(result.content).toMatch(/keep helping/i);
     expect(result.content).not.toMatch(/name and URL/);
-    // CL-7141: no fake `/plugins?connect=mcp` deep link — there is no
+    // no fake `/plugins?connect=mcp` deep link — there is no
     // generic add-custom-MCP-server card to land on, so the fallback
     // points at the Plugins page's own connector list in plain prose.
     expect(result.content).not.toContain("/plugins?connect=mcp");
diff --git a/tools/connections/src/tool.ts b/tools/connections/src/tool.ts
index b9ea959c51..07ccb87ee8 100644
--- a/tools/connections/src/tool.ts
+++ b/tools/connections/src/tool.ts
@@ -14,7 +14,7 @@
 // names and links. MCP servers are not a stock Interchange concept, so a
 // tenant-minted `mcp:<slug>` connection is no longer listed: a curated
 // preset is still offered as a card, but the tool cannot say whether one is
-// already connected (CL-8164 moves MCP server config to the connections
+// already connected (moves MCP server config to the connections
 // library).
 //
 // Approval: `list_connections` reads only, so it declares no `approval`
@@ -85,7 +85,7 @@ function clientConfig(env: WorkflowConnectionEnv) {
 
 /** `/plugins?connect=<connectorId>` — a deep link into the Plugins
  * panel that auto-opens the named connector's own connect card
- * (CL-7141: `apps/web/src/pages/plugins-page.tsx` reads this `connect`
+ * (`apps/web/src/pages/plugins-page.tsx` reads this `connect`
  * query param on mount and hands it to the same
  * `requestPluginsConnect` path an in-app "Fix it" click uses). */
 function connectDeepLink(connectorId: string): string {
@@ -93,7 +93,7 @@ function connectDeepLink(connectorId: string): string {
 }
 
 /** There is no fixed id — and no generic add-custom-MCP-server form at
- * all — to deep-link to (CL-7141; see
+ * all — to deep-link to (; see
  * `packages/plugins-ui/src/mcp-servers-section.tsx`'s header for why:
  * only curated presets are self-serve installable, an MCP server has
  * no id before it's tenant-minted at connect time). So this stays
@@ -102,10 +102,10 @@ function connectDeepLink(connectorId: string): string {
 const ADD_MCP_SERVER_GUIDANCE =
   "point them to the Plugins page in this workbench — every connector it can add lives there";
 
-/** `/plugins?connect=mcp:<slug>` — a curated preset's own card
- * (CL-6152). Presets are still tenant-minted `mcp:<slug>` connections
- * once connected, but the *card* a human clicks to start one is fixed
- * and known ahead of time, unlike a hand-typed custom MCP server. */
+/** `/plugins?connect=mcp:<slug>` — a curated preset's own card. Presets
+ * are still tenant-minted `mcp:<slug>` connections once connected, but
+ * the *card* a human clicks to start one is fixed and known ahead of
+ * time, unlike a hand-typed custom MCP server. */
 function presetDeepLink(slug: string): string {
   return `/plugins?connect=mcp:${slug}`;
 }
diff --git a/tools/github/README.md b/tools/github/README.md
index 3d4dcf79b3..978e9a3fc9 100644
--- a/tools/github/README.md
+++ b/tools/github/README.md
@@ -1,7 +1,7 @@
 # @corbits/github-tools
 
 A minimal GitHub integration: search, plus the pull-request reads and
-one write a code review needs — two `@intx/agent` tool bundles. Built for the last-30-days-research workflow port (CL-5997),
+one write a code review needs — two `@intx/agent` tool bundles. Built for the last-30-days-research workflow port,
 mirroring the OG gtm-workbench's `packages/tools-github` client against
 GitHub's real public REST search API.
 
diff --git a/tools/github/src/manifest.test.ts b/tools/github/src/manifest.test.ts
index 9e7fb89467..d2e73db3d5 100644
--- a/tools/github/src/manifest.test.ts
+++ b/tools/github/src/manifest.test.ts
@@ -2,7 +2,7 @@
 // `PackageJSON` manifest schema (`@intx/types/src/package-json.ts`)
 // with an `interchange.credentials` declaration the deploy-time
 // `ToolPackageManifest` harvests into `topLevel[].credentials` — the
-// static half of the credential-binding seam CL-6028 adopts. GitHub's
+// static half of the credential-binding seam adopts. GitHub's
 // credential is optional at runtime (an unresolved "github" handle
 // degrades to unauthenticated calls, never "not connected" — see
 // `tool.ts`), but the manifest still declares the handle that *could*
diff --git a/tools/github/src/repos.ts b/tools/github/src/repos.ts
index 0f46d6ca68..5af2954290 100644
--- a/tools/github/src/repos.ts
+++ b/tools/github/src/repos.ts
@@ -1,11 +1,11 @@
-// The connect-github card's repo picker (CL-6345): the repositories the
+// The connect-github card's repo picker: the repositories the
 // connected token can reach, most recently pushed first.
 //
 // One call, on purpose. This used to decorate every row with an exact
 // open-PR count from the search API, which cost one `/search/issues` call
 // per repo fired concurrently — and GitHub's search API allows 30 requests
 // a minute and secondary-rate-limits bursts, so any account past a handful
-// of repos got a 403 and the whole picker failed to load (CL-7189). The
+// of repos got a 403 and the whole picker failed to load. The
 // count was decoration on a checkbox row; the push timestamp that explains
 // the ordering rides along on the list response for free.
 import { type } from "arktype";
@@ -83,7 +83,7 @@ const GitHubAuthenticatedUser = type({ login: "string" });
 /**
  * The authenticated PAT's own login — this connector authenticates a
  * person, not an organization (GitHub App/OAuth org-level connect is
- * CL-6343, not built here), so a caller showing "connected as" shows
+ *, not built here), so a caller showing "connected as" shows
  * this login, not an org name.
  */
 export async function fetchAuthenticatedLogin(config: GitHubClientConfig): Promise<string> {
diff --git a/tools/granola/README.md b/tools/granola/README.md
index 0f544cceed..a8ae002360 100644
--- a/tools/granola/README.md
+++ b/tools/granola/README.md
@@ -1,8 +1,8 @@
 # @corbits/granola-tools
 
 A minimal Granola integration: two client calls, one `@intx/agent` tool
-bundle. Built for the morning-brief workflow (CL-5993); `granola_get_note`
-was added by the pain-point-collateral workflow (CL-5995) — neither is
+bundle. Built for the morning-brief workflow; `granola_get_note`
+was added by the pain-point-collateral workflow — neither is
 specific to its origin workflow, so both stay in this one package rather
 than forking a second Granola client.
 
@@ -25,7 +25,7 @@ content naming the source "not connected", so a calling agent can
 degrade gracefully instead of failing its turn.
 
 This package declares one credential handle, `granola`, in its
-`package.json`'s `interchange.credentials` field (CL-6028). Connect
+`package.json`'s `interchange.credentials` field. Connect
 Granola once, in Settings · Connections, and every workflow that pins
 `@corbits/granola-tools` and binds this handle — `granola-call`,
 `process-granola-call`, `pain-point-collateral`, `collateral-generation`,
@@ -36,7 +36,7 @@ Launch-time resolution (`buildCredentialDelivery`) is proven in
 `test/credential-delivery.drizzle.test.ts`; the full chain — seeded
 credential through the sidecar's step wiring
 (`apps/sidecar/src/step-agent-tools.ts`) to this bundle's tool call — is
-proven in `test/credential-wiring-e2e.drizzle.test.ts` (CL-6032). See
+proven in `test/credential-wiring-e2e.drizzle.test.ts`. See
 `docs/credential-wiring.md` for the end-to-end picture and the provider
 plugins involved (this package's Granola credential uses the vendored
 `http` / Bearer plugin; `@corbits/linear-tools`'s Linear credential does
diff --git a/tools/granola/src/manifest.test.ts b/tools/granola/src/manifest.test.ts
index d308a98d0a..5a685c798e 100644
--- a/tools/granola/src/manifest.test.ts
+++ b/tools/granola/src/manifest.test.ts
@@ -2,7 +2,7 @@
 // `PackageJSON` manifest schema (`@intx/types/src/package-json.ts`)
 // with an `interchange.credentials` declaration the deploy-time
 // `ToolPackageManifest` harvests into `topLevel[].credentials` — the
-// static half of the credential-binding seam CL-6028 adopts. Not a test
+// static half of the credential-binding seam adopts. Not a test
 // of the vendored schema itself, only that this package's manifest is
 // shaped the way that schema requires.
 import { expect, test } from "bun:test";
diff --git a/tools/granola/src/tool.ts b/tools/granola/src/tool.ts
index 65326684be..97d523732e 100644
--- a/tools/granola/src/tool.ts
+++ b/tools/granola/src/tool.ts
@@ -6,8 +6,8 @@
 // and say so honestly, never to have the run itself fail because one
 // source is unreachable.
 //
-// CL-6028 declared this package's "granola" credential handle in
-// `package.json` (`interchange.credentials`); CL-6032 wires the runtime
+// declared this package's "granola" credential handle in
+// `package.json` (`interchange.credentials`); wires the runtime
 // half. `env.credentials` is the harness's consumer-gated capability
 // (`@intx/harness/src/credential-capability.ts`,
 // `createCredentialCapability`): the sidecar's step-invoker binding
diff --git a/tools/granola/test/credential-delivery.drizzle.test.ts b/tools/granola/test/credential-delivery.drizzle.test.ts
index 6f97c3b195..30963bdf73 100644
--- a/tools/granola/test/credential-delivery.drizzle.test.ts
+++ b/tools/granola/test/credential-delivery.drizzle.test.ts
@@ -4,7 +4,7 @@
 // its own Postgres schema, never the developer's or the walking-skeleton
 // suite's.
 //
-// Proves the launch-time half of CL-6028's credential-binding adoption
+// Proves the launch-time half of its credential-binding adoption
 // using only the platform's own, already-built functions
 // (`buildCredentialDelivery`, `createEnvKeyCredentialCipher`,
 // `credentialAad`) — never a hand-rolled resolver:
diff --git a/tools/interaction/src/client.ts b/tools/interaction/src/client.ts
index 6d07594cdd..a9599b4a7a 100644
--- a/tools/interaction/src/client.ts
+++ b/tools/interaction/src/client.ts
@@ -24,7 +24,7 @@ export interface AskUserQuestion {
    * Stable id for this card. When omitted, `postQuestion` mints one.
    * `ask_user` always supplies `questionIdForCall(call.id)` so a retry of
    * the same tool call re-posts the same id — the participants/messages
-   * route treats that as a no-op and returns the existing card (CL-7248).
+   * route treats that as a no-op and returns the existing card.
    */
   readonly questionId?: string;
 }
diff --git a/tools/interaction/src/tool.ts b/tools/interaction/src/tool.ts
index 4317cc997a..a9b4c83598 100644
--- a/tools/interaction/src/tool.ts
+++ b/tools/interaction/src/tool.ts
@@ -59,7 +59,7 @@ function clientConfig(env: AskUserEnv): AskUserClientConfig {
  * `postQuestion` stamps `questionIdForCall(call.id)` (derived from the
  * tool-call id, not minted per attempt) on the outbound question card, so a
  * crash-retry of the same call re-posts the same id and the write path
- * returns the existing card rather than a duplicate (CL-7248).
+ * returns the existing card rather than a duplicate.
  */
 async function runAskUser(env: AskUserEnv, call: ToolCall): Promise<ToolResult> {
   const parsed = AskUserInput(call.arguments);
diff --git a/tools/linear/README.md b/tools/linear/README.md
index 478af52c48..b083e3a6e9 100644
--- a/tools/linear/README.md
+++ b/tools/linear/README.md
@@ -1,7 +1,7 @@
 # @corbits/linear-tools
 
 A minimal Linear integration: one client call, one `@intx/agent` tool
-bundle. Built for the morning-brief workflow (CL-5993) but not specific
+bundle. Built for the morning-brief workflow but not specific
 to it — any agent that needs a user's recently updated Linear issues
 pins this package.
 
@@ -23,7 +23,7 @@ content naming the source "not connected", so a calling agent can
 degrade gracefully instead of failing its turn.
 
 This package declares one credential handle, `linear`, in its
-`package.json`'s `interchange.credentials` field (CL-6028). Connect
+`package.json`'s `interchange.credentials` field. Connect
 Linear once, in Settings · Connections, and every workflow that pins
 `@corbits/linear-tools` and binds this handle — `collateral-generation`,
 `morning-brief` — resolves the same tenant-owned credential at launch;
@@ -33,7 +33,7 @@ Launch-time resolution (`buildCredentialDelivery`) is proven in
 `test/credential-delivery.drizzle.test.ts`; the full chain — seeded
 credential through the sidecar's step wiring
 (`apps/sidecar/src/step-agent-tools.ts`) to this bundle's tool call — is
-proven in `test/credential-wiring-e2e.drizzle.test.ts` (CL-6032).
+proven in `test/credential-wiring-e2e.drizzle.test.ts`.
 
 **Provider plugin.** Linear's API expects the raw key verbatim in
 `authorization`, not a `Bearer `-prefixed token. `@intx/harness`'s
diff --git a/tools/linear/src/manifest.test.ts b/tools/linear/src/manifest.test.ts
index f09976b744..af9fe641e8 100644
--- a/tools/linear/src/manifest.test.ts
+++ b/tools/linear/src/manifest.test.ts
@@ -2,7 +2,7 @@
 // `PackageJSON` manifest schema (`@intx/types/src/package-json.ts`)
 // with an `interchange.credentials` declaration the deploy-time
 // `ToolPackageManifest` harvests into `topLevel[].credentials` — the
-// static half of the credential-binding seam CL-6028 adopts. Not a test
+// static half of the credential-binding seam adopts. Not a test
 // of the vendored schema itself, only that this package's manifest is
 // shaped the way that schema requires.
 import { expect, test } from "bun:test";
diff --git a/tools/linear/src/tool.ts b/tools/linear/src/tool.ts
index b44b8598ec..60ba6fc6c0 100644
--- a/tools/linear/src/tool.ts
+++ b/tools/linear/src/tool.ts
@@ -6,8 +6,8 @@
 // and say so honestly, never to have the run itself fail because one
 // source is unreachable.
 //
-// CL-6028 declared this package's "linear" credential handle in
-// `package.json` (`interchange.credentials`); CL-6032 wires the runtime
+// declared this package's "linear" credential handle in
+// `package.json` (`interchange.credentials`); wires the runtime
 // half. `env.credentials` is the harness's consumer-gated capability
 // (`@intx/harness/src/credential-capability.ts`,
 // `createCredentialCapability`): the sidecar's step-invoker binding
diff --git a/tools/linear/test/credential-delivery.drizzle.test.ts b/tools/linear/test/credential-delivery.drizzle.test.ts
index ff2068baa1..a659f2d184 100644
--- a/tools/linear/test/credential-delivery.drizzle.test.ts
+++ b/tools/linear/test/credential-delivery.drizzle.test.ts
@@ -4,7 +4,7 @@
 // its own Postgres schema, never the developer's or the walking-skeleton
 // suite's.
 //
-// Proves the launch-time half of CL-6028's credential-binding adoption
+// Proves the launch-time half of its credential-binding adoption
 // using only the platform's own, already-built functions
 // (`buildCredentialDelivery`, `createEnvKeyCredentialCipher`,
 // `credentialAad`) — never a hand-rolled resolver:
diff --git a/tools/linear/test/credential-wiring-e2e.drizzle.test.ts b/tools/linear/test/credential-wiring-e2e.drizzle.test.ts
index e718be2092..8d0c3ab784 100644
--- a/tools/linear/test/credential-wiring-e2e.drizzle.test.ts
+++ b/tools/linear/test/credential-wiring-e2e.drizzle.test.ts
@@ -2,7 +2,7 @@
 // `credential-delivery.drizzle.test.ts` (its own schema, never the
 // developer's or the walking-skeleton suite's) and
 // `@corbits/granola-tools`'s `test/credential-wiring-e2e.drizzle.test.ts`
-// (CL-6032), which this test follows step for step.
+//, which this test follows step for step.
 //
 // Proves the FULL "connect once, workflows just work" chain end to end
 // through the platform's own, already-built functions -- never a
diff --git a/tools/mcp/src/allowance.ts b/tools/mcp/src/allowance.ts
index 2fb9d575a1..8a5084a03e 100644
--- a/tools/mcp/src/allowance.ts
+++ b/tools/mcp/src/allowance.ts
@@ -1,4 +1,4 @@
-// CL-6345: `mcp_call`'s declarative grant-allowance classification. The
+// `mcp_call`'s declarative grant-allowance classification. The
 // tool itself stays `approval: "ask"` (the static floor cannot vary by
 // downstream tool — see `./tool.ts`'s header), so every call still
 // parks; the hub's allowance gate then consults this classifier to
diff --git a/tools/mcp/src/tool.ts b/tools/mcp/src/tool.ts
index d537db8a34..6a5ac64318 100644
--- a/tools/mcp/src/tool.ts
+++ b/tools/mcp/src/tool.ts
@@ -39,7 +39,7 @@
 // call sees whether the tool the model is about to invoke claims to
 // be read-only — an honest signal, not an enforcement mechanism. This
 // is the [Intx gap] the task named up front (no dynamic/late-bound
-// approval floor); "ask always" stays the static floor. CL-6345 layers
+// approval floor); "ask always" stays the static floor. layers
 // the grant allowance on top at the hub: a parked `mcp_call` whose
 // downstream tool the server itself marks `readOnlyHint: true` (see
 // `./allowance.ts`'s classifier, which re-verifies the claim live) and
diff --git a/tools/reddit/README.md b/tools/reddit/README.md
index f69158a332..d9c603ba63 100644
--- a/tools/reddit/README.md
+++ b/tools/reddit/README.md
@@ -1,7 +1,7 @@
 # @corbits/reddit-tools
 
 A minimal Reddit search integration: one client, one `@intx/agent` tool
-bundle. Built for the reddit-opportunity-scanner workflow (CL-5994) but
+bundle. Built for the reddit-opportunity-scanner workflow but
 not specific to it — any agent that needs to search Reddit pins this
 package.
 
diff --git a/tools/reddit/src/manifest.test.ts b/tools/reddit/src/manifest.test.ts
index 850b6553a1..85e80c922e 100644
--- a/tools/reddit/src/manifest.test.ts
+++ b/tools/reddit/src/manifest.test.ts
@@ -2,7 +2,7 @@
 // `PackageJSON` manifest schema (`@intx/types/src/package-json.ts`)
 // with an `interchange.credentials` declaration the deploy-time
 // `ToolPackageManifest` harvests into `topLevel[].credentials` — the
-// static half of the credential-binding seam CL-6028 adopts. Not a test
+// static half of the credential-binding seam adopts. Not a test
 // of the vendored schema itself, only that this package's manifest is
 // shaped the way that schema requires.
 import { expect, test } from "bun:test";
diff --git a/tools/skills-tools/src/client.test.ts b/tools/skills-tools/src/client.test.ts
index 24cd199dc2..c548c60c13 100644
--- a/tools/skills-tools/src/client.test.ts
+++ b/tools/skills-tools/src/client.test.ts
@@ -19,7 +19,7 @@ function testConfig(fetchImpl: typeof fetch): SkillsToolClientConfig {
   };
 }
 
-// CL-8086: the workflow-skills HTTP surface these four calls used to
+// the workflow-skills HTTP surface these four calls used to
 // reach was deleted, and no stock Interchange route yet serves skill
 // content, so each fails closed with an explicit error instead of
 // reaching a dead route.
diff --git a/tools/skills-tools/src/client.ts b/tools/skills-tools/src/client.ts
index 2337683808..ce488db35b 100644
--- a/tools/skills-tools/src/client.ts
+++ b/tools/skills-tools/src/client.ts
@@ -1,7 +1,7 @@
 // A minimal client for the workflow-run-authenticated surfaces
 // `@corbits/skills-tools`' tools call.
 //
-// CL-8086: `@corbits/skills`' `createWorkflowSkillRoutes` (formerly
+// `@corbits/skills`' `createWorkflowSkillRoutes` (formerly
 // mounted at `/api/workflow-skills`, serving `list`/`search`/`load` plus
 // the `create`/`update` writes) was deleted — skills are native
 // `kind:"skill"` hub assets now, and no stock Interchange route yet lets
@@ -91,12 +91,12 @@ function skillPinEndpoint(config: SkillsToolClientConfig, path: string): string
 }
 
 const NO_STOCK_SKILL_CONTENT_ROUTE =
-  "Skill content has no stock Interchange HTTP route (CL-8086): the " +
+  "Skill content has no stock Interchange HTTP route: the " +
   "workbench-specific skills registry that used to serve it was removed, " +
   "and no replacement has been added to @intx/hub-api yet.";
 
 /** Every skill this run can see, index-only (name + description, no
- * body). No stock Interchange route serves this yet (CL-8086); fails
+ * body). No stock Interchange route serves this yet; fails
  * closed rather than reaching a dead route or reading as an empty
  * registry. */
 export function listSkills(_config: SkillsToolClientConfig): Promise<readonly SkillIndexEntry[]> {
@@ -104,14 +104,14 @@ export function listSkills(_config: SkillsToolClientConfig): Promise<readonly Sk
 }
 
 /** Loads one skill's full body by name. No stock Interchange route
- * serves this yet (CL-8086); fails closed rather than fabricating
+ * serves this yet; fails closed rather than fabricating
  * content. */
 export function loadSkill(_config: SkillsToolClientConfig, _name: string): Promise<SkillDetail> {
   return Promise.reject(new Error(NO_STOCK_SKILL_CONTENT_ROUTE));
 }
 
 /** Creates a new, always tenant-scoped skill. No stock Interchange route
- * accepts skill content yet (CL-8086); fails closed rather than
+ * accepts skill content yet; fails closed rather than
  * fabricating success. */
 export function createSkill(
   _config: SkillsToolClientConfig,
@@ -125,7 +125,7 @@ export function createSkill(
 }
 
 /** Republishes an existing skill's body. No stock Interchange route
- * accepts skill content yet (CL-8086); fails closed rather than
+ * accepts skill content yet; fails closed rather than
  * fabricating success. */
 export function updateSkill(
   _config: SkillsToolClientConfig,
diff --git a/tools/skills-tools/src/tool.test.ts b/tools/skills-tools/src/tool.test.ts
index a124789149..ac55592c36 100644
--- a/tools/skills-tools/src/tool.test.ts
+++ b/tools/skills-tools/src/tool.test.ts
@@ -71,7 +71,7 @@ test("read_skill rejects a call missing the name without calling out", async ()
   expect(result.content).toMatch(/invalid input/);
 });
 
-// CL-8086: the workflow-skills HTTP surface these tools used to call was
+// the workflow-skills HTTP surface these tools used to call was
 // deleted, and no stock Interchange route yet serves skill content, so
 // each fails closed with an explicit error instead of a fabricated
 // result.
diff --git a/tools/skills-tools/src/tool.ts b/tools/skills-tools/src/tool.ts
index 67a4f46774..2691c1f126 100644
--- a/tools/skills-tools/src/tool.ts
+++ b/tools/skills-tools/src/tool.ts
@@ -6,7 +6,7 @@
 // `@corbits/capability-tools`' `request_capability`, over the same
 // workflow-run-authenticated surfaces `./client.ts` calls.
 //
-// `create_skill` and `update_skill` (CL-6209) grant no credentials or
+// `create_skill` and `update_skill` grant no credentials or
 // capability pins and write only a tenant-internal skill row — a
 // skill's body is durable know-how, not something any agent's
 // capabilities or autonomy hinge on until a human separately pins it
diff --git a/tools/tools-skills/src/client.test.ts b/tools/tools-skills/src/client.test.ts
index 1e21e1b267..4c10116830 100644
--- a/tools/tools-skills/src/client.test.ts
+++ b/tools/tools-skills/src/client.test.ts
@@ -8,7 +8,7 @@ const CONFIG = {
   runAddress: "run@runs.example",
 };
 
-// CL-8086: the workflow-skills HTTP surface these calls used to reach was
+// the workflow-skills HTTP surface these calls used to reach was
 // deleted, and no stock Interchange route yet serves skill content. Every
 // call fails closed with an explicit error rather than degrading to an
 // empty registry.
diff --git a/tools/tools-skills/src/client.ts b/tools/tools-skills/src/client.ts
index 0c5427d30b..5aefaf1ade 100644
--- a/tools/tools-skills/src/client.ts
+++ b/tools/tools-skills/src/client.ts
@@ -1,4 +1,4 @@
-// CL-8086: the Workbench skill registry (`@corbits/skills`' HTTP surface
+// the Workbench skill registry (`@corbits/skills`' HTTP surface
 // this client used to call, mounted at `/api/workflow-skills`) was
 // deleted — skills are native `kind:"skill"` hub assets now, and there is
 // no stock Interchange route yet that lets a workflow-run bearer identity
@@ -23,7 +23,7 @@ export type SkillIndexItem = {
 export type LoadedSkill = SkillIndexItem & { readonly body: string };
 
 const NO_STOCK_SKILL_CONTENT_ROUTE =
-  "Skill content has no stock Interchange HTTP route (CL-8086): the " +
+  "Skill content has no stock Interchange HTTP route: the " +
   "workbench-specific skills registry that used to serve it was removed, " +
   "and no replacement has been added to @intx/hub-api yet.";
 
diff --git a/tools/tools-skills/src/tool.test.ts b/tools/tools-skills/src/tool.test.ts
index 2029c16c6c..9f947d3581 100644
--- a/tools/tools-skills/src/tool.test.ts
+++ b/tools/tools-skills/src/tool.test.ts
@@ -42,7 +42,7 @@ test("no tool's input schema accepts a tenant or principal argument", () => {
   }
 });
 
-// CL-8086: the registry HTTP surface this client used to call was
+// the registry HTTP surface this client used to call was
 // deleted, and no stock Interchange route yet serves skill content, so
 // every call fails closed rather than reaching a live registry.
 test("skills_list surfaces the missing stock route as an error, never an empty list", async () => {
diff --git a/tools/web-search/README.md b/tools/web-search/README.md
index 90cda2c984..1b1f5e0846 100644
--- a/tools/web-search/README.md
+++ b/tools/web-search/README.md
@@ -1,7 +1,7 @@
 # @corbits/web-search-tools
 
 A minimal web-search integration: one client call, one `@intx/agent` tool
-bundle. Built for the last-30-days-research workflow port (CL-5997);
+bundle. Built for the last-30-days-research workflow port;
 backed by Exa, the same provider the OG gtm-workbench's
 `last30days-research` workflow called for this source
 (`packages/tools-exa`) — a real, honest backend rather than a placeholder.
diff --git a/tools/web-search/src/manifest.test.ts b/tools/web-search/src/manifest.test.ts
index 9128b32cf2..08f19fbb51 100644
--- a/tools/web-search/src/manifest.test.ts
+++ b/tools/web-search/src/manifest.test.ts
@@ -2,7 +2,7 @@
 // `PackageJSON` manifest schema (`@intx/types/src/package-json.ts`)
 // with an `interchange.credentials` declaration the deploy-time
 // `ToolPackageManifest` harvests into `topLevel[].credentials` — the
-// static half of the credential-binding seam CL-6028 adopts. Not a test
+// static half of the credential-binding seam adopts. Not a test
 // of the vendored schema itself, only that this package's manifest is
 // shaped the way that schema requires.
 import { expect, test } from "bun:test";
diff --git a/tools/workflow-authoring/README.md b/tools/workflow-authoring/README.md
index 9e7e0bcfc6..849cfa6151 100644
--- a/tools/workflow-authoring/README.md
+++ b/tools/workflow-authoring/README.md
@@ -1,7 +1,7 @@
 # @corbits/workflow-authoring-tools
 
 The `@intx/agent` tool bundle over `@corbits/workflows`'s `./authoring`'s
-workflow-run-authenticated routes (CL-7360): an agent writes an ordinary
+workflow-run-authenticated routes: an agent writes an ordinary
 workflow code package into a `kind: "workflow"` hub asset, republishes it,
 and reads it back. See [docs/workflow-source-authoring.md](../../docs/workflow-source-authoring.md)
 for the contract this implements.
@@ -49,7 +49,7 @@ on.
 | `wf_deploy_preview`    | `POST /api/workflow-workflow-authoring/:assetId/deploy/preview` | `previewDeployWorkflow` |
 | `workflow_deploy`      | `POST /api/tenants/:tenantId/workflows/deployments` (stock)     | `deployWorkflow`        |
 
-`workflow_deploy` speaks stock Interchange routes (CL-8171): it reads the
+`workflow_deploy` speaks stock Interchange routes: it reads the
 tenant's resolved inference catalog from `GET /api/tenants/:tenantId/models`,
 flattens it into one priority-ordered `sourceOfferingIds` chain, and posts an
 `asset`/`source` deploy naming the commit. The run bearer authenticates both.
@@ -59,7 +59,7 @@ The four authoring operations still call the Workbench-specific
 smart-HTTP on the asset repo, which no run-bearer credential reaches today:
 `createGitTokenAuth` accepts only an `itx_pat_`/`itx_svc_` git token, and the
 stock mint route (`POST /api/tenants/:tenantId/git-tokens`) refuses a caller
-with no browser session. CL-8171 carries that upstream ask.
+with no browser session. carries that upstream ask.
 
 A hub refusal surfaces as `WorkflowAuthoringRequestError` (`status`, `code`,
 `currentHeadSha` on a conflict); the bundle lets it throw, and
diff --git a/tools/workflow-authoring/src/client.ts b/tools/workflow-authoring/src/client.ts
index aeb33c51d8..45408dfb88 100644
--- a/tools/workflow-authoring/src/client.ts
+++ b/tools/workflow-authoring/src/client.ts
@@ -16,7 +16,7 @@
 // smart-HTTP on the asset repo, which no run-bearer credential can reach
 // today: `createGitTokenAuth` accepts only an `itx_pat_`/`itx_svc_` git
 // token, and the stock mint route refuses a caller with no browser
-// session. See CL-8171 for the upstream ask that would let these four
+// session. See for the upstream ask that would let these four
 // follow `deploy` onto stock routes.
 import { type } from "arktype";
 import { runBearerHeaders, runBearerFetch, type RunBearerClientConfig } from "./run-bearer";
diff --git a/tools/workflow-authoring/src/tool.ts b/tools/workflow-authoring/src/tool.ts
index 7518684f13..fc2b0c902e 100644
--- a/tools/workflow-authoring/src/tool.ts
+++ b/tools/workflow-authoring/src/tool.ts
@@ -6,8 +6,7 @@
 // source is not a side effect (docs/workflow-model.md, "Authority
 // boundaries"). `workflow_deploy` does — deploying is what makes a
 // workflow selectable as a routine target, so a human sees the deploy
-// intent and approves it before the tool call ever reaches the hub (CL-7362
-// still owns showing the probed capability surface on that approval card;
+// intent and approves it before the tool call ever reaches the hub (// still owns showing the probed capability surface on that approval card;
 // today's snapshot is the tool call's own arguments).
 //
 // A thrown error here is the honest result: `@intx/agent`'s tool runner
@@ -325,7 +324,7 @@ export const workflowAuthoringTools = defineTool<WorkflowAuthoringEnv>({
           "so the approval reflects the real committed source, not just " +
           "an asset id. Grants/capabilities are stamped by the native " +
           "install+probe+gate this call runs, which the human does not " +
-          "see a preview of yet (CL-7362); say so if you explain this " +
+          "see a preview of yet; say so if you explain this " +
           "approval to a human. Inference sources come from the " +
           "workbench's own catalog — never pass a model or credential.",
         inputSchema: {