From 366ee1d0fc2c920c583915ec504c921674bda9a7 Mon Sep 17 00:00:00 2001
From: Sorra the Orc <ross@example.com>
Date: Wed, 8 Jul 2026 14:07:18 +0100
Subject: [PATCH 1/8] Add model-display Pi extension (LP-0MR4ZIGDT004A3E1)

Adds a Pi extension that shows the resolved provider/model from
llama-proxy responses in Pi's status bar. Reads the X-Resolved-Model
header added by the proxy and displays it alongside the selected
model alias in the format: <selected-model> -> <provider>/<model-id>

Registered alongside the existing Worklog extension in package.json.
---
 packages/tui/extensions/model-display.ts | 61 ++++++++++++++++++++++++
 packages/tui/extensions/package.json     |  3 +-
 2 files changed, 63 insertions(+), 1 deletion(-)
 create mode 100644 packages/tui/extensions/model-display.ts

diff --git a/packages/tui/extensions/model-display.ts b/packages/tui/extensions/model-display.ts
new file mode 100644
index 00000000..4b3a61f8
--- /dev/null
+++ b/packages/tui/extensions/model-display.ts
@@ -0,0 +1,61 @@
+/**
+ * Model Display Extension (LP-0MR4ZIGDT004A3E1)
+ *
+ * Displays the resolved provider/model from llama-proxy responses in Pi's
+ * status bar.
+ *
+ * Shows the format: `<selected-model> → <provider>/<model-id>`
+ *
+ * - Listens to `after_provider_response` to read the `X-Resolved-Model` header
+ *   that the llama-proxy adds to every proxy-routed request.
+ * - Listens to `model_select` to track the currently selected Pi model alias.
+ * - Updates the status bar on every request so the user always sees which
+ *   provider and model actually served their request.
+ *
+ * Requires the llama-proxy to be configured as Pi's provider and to include
+ * the `X-Resolved-Model` header in its responses (see LP-0MR4ZIGDT004A3E1).
+ */
+import type { ExtensionAPI } from "@earendil-works/pi-coding-agent";
+
+export default function (pi: ExtensionAPI) {
+  // Track the currently selected model alias (e.g., "plan", "code")
+  let selectedModel: string | null = null;
+
+  // Track the resolved model from the last proxy response
+  let resolvedModel: string | null = null;
+
+  // Update the status display
+  function updateStatus(ctx: { ui: { setStatus: (key: string, value: string) => void } }) {
+    if (selectedModel && resolvedModel) {
+      ctx.ui.setStatus("model-display", `${selectedModel} → ${resolvedModel}`);
+    } else if (selectedModel) {
+      ctx.ui.setStatus("model-display", `${selectedModel} → (pending)`);
+    } else if (resolvedModel) {
+      ctx.ui.setStatus("model-display", `→ ${resolvedModel}`);
+    }
+  }
+
+  // Track model selection changes
+  pi.on("model_select", (_event, ctx) => {
+    // event.model has provider and id fields
+    const model = _event.model;
+    if (model && model.id) {
+      selectedModel = model.id;
+    }
+    updateStatus(ctx);
+  });
+
+  // Extract resolved model from proxy response headers
+  pi.on("after_provider_response", (event, ctx) => {
+    // event.headers contains the normalized response headers
+    // X-Resolved-Model header is lowercase in normalized form
+    const headers = event.headers as Record<string, string>;
+    const headerValue: string | undefined =
+      headers["x-resolved-model"] ?? headers["X-Resolved-Model"];
+
+    if (headerValue) {
+      resolvedModel = headerValue;
+    }
+    updateStatus(ctx);
+  });
+}
diff --git a/packages/tui/extensions/package.json b/packages/tui/extensions/package.json
index dded3f5d..fc32cdb4 100644
--- a/packages/tui/extensions/package.json
+++ b/packages/tui/extensions/package.json
@@ -6,7 +6,8 @@
   "description": "Worklog Pi TUI extension — Work item browser for Pi coding agent",
   "pi": {
     "extensions": [
-      "./Worklog/index.ts"
+      "./Worklog/index.ts",
+      "./model-display.ts"
     ]
   }
 }

From 29a15d5e1948f56569c23d4c016c4e3f36182769 Mon Sep 17 00:00:00 2001
From: Sorra the Orc <>
Date: Wed, 8 Jul 2026 14:22:36 +0100
Subject: [PATCH 2/8] WL-0MRC3F30E008XJN3: Integrate model-display extension
 into the Worklog extension

- Create Worklog/model-display.ts with moved logic and status key renamed
  from 'model-display' to 'worklog-model'
- Import and register registerModelDisplay() in Worklog/index.ts
- Delete packages/tui/extensions/model-display.ts (old standalone extension)
- Update package.json to remove ./model-display.ts from extensions list
- Add unit tests (15 tests) for registerModelDisplay() covering:
  - model_select and after_provider_response event registration
  - Status bar updates with correct formatting
  - Fallback states (pending model, no model)
  - Multiple event interactions
  - Header key case sensitivity
---
 packages/tui/extensions/Worklog/index.ts      |   2 +
 .../extensions/Worklog/model-display.test.ts  | 238 ++++++++++++++++++
 .../extensions/{ => Worklog}/model-display.ts |  39 ++-
 packages/tui/extensions/package.json          |   3 +-
 4 files changed, 270 insertions(+), 12 deletions(-)
 create mode 100644 packages/tui/extensions/Worklog/model-display.test.ts
 rename packages/tui/extensions/{ => Worklog}/model-display.ts (56%)

diff --git a/packages/tui/extensions/Worklog/index.ts b/packages/tui/extensions/Worklog/index.ts
index 61997898..f852a805 100644
--- a/packages/tui/extensions/Worklog/index.ts
+++ b/packages/tui/extensions/Worklog/index.ts
@@ -19,6 +19,7 @@ import { worklogConfig } from './config.js';
 import { reloadSettings, currentSettings, STAGE_MAP, VALID_STAGES, updateSettings, openSettingsOverlay } from './lib/settings.js';
 import { runWl, defaultListWorkItems, defaultListWorkItemsWithStage, createDefaultListWorkItems, createListWorkItemsWithStage, createDefaultListWorkItemsDb, createListWorkItemsWithStageDb, fetchTotalActionableCountDb } from './lib/tools.js';
 import { registerAutoInject } from './lib/auto-inject.js';
+import { registerModelDisplay } from './model-display.js';
 import { INSTALL_GUARDRAILS } from './lib/guardrails.js';
 import { registerSkillPathTool } from './lib/skill-path.js';
 import { registerRecoveryModule } from './lib/recovery/register-recovery.js';
@@ -88,6 +89,7 @@ export function createWorklogBrowseExtension(deps: WorklogBrowseDependencies = {
   return function registerWorklogBrowseExtension(pi: ExtensionAPI): void {
     registerActivityIndicator(pi, () => currentSettings.showActivityIndicator);
     registerAutoInject(pi);
+    registerModelDisplay(pi);
     INSTALL_GUARDRAILS(pi, { enabled: currentSettings.guardrailsEnabled });
 
     // ── Skill path discovery tool ─────────────────────────────────
diff --git a/packages/tui/extensions/Worklog/model-display.test.ts b/packages/tui/extensions/Worklog/model-display.test.ts
new file mode 100644
index 00000000..a502a174
--- /dev/null
+++ b/packages/tui/extensions/Worklog/model-display.test.ts
@@ -0,0 +1,238 @@
+/**
+ * Unit tests for model-display.ts — Model/provider display in Pi's status bar.
+ *
+ * Verifies that:
+ * 1. registerModelDisplay() sets up event listeners for model_select and
+ *    after_provider_response.
+ * 2. model_select updates the status bar with the selected model alias.
+ * 3. after_provider_response updates the status bar with the resolved
+ *    provider/model from X-Resolved-Model header.
+ * 4. The status bar key is "worklog-model" (not "model-display").
+ * 5. Fallback display states work correctly (pending model, no model).
+ * 6. The module exports the MODEL_DISPLAY_STATUS_KEY constant.
+ *
+ * Run: npx vitest run packages/tui/extensions/Worklog/model-display.test.ts
+ */
+
+import { describe, it, expect, vi, beforeEach } from 'vitest';
+
+// ── Mock @earendil-works/pi-coding-agent ──────────────────────────────
+
+vi.mock('@earendil-works/pi-coding-agent', () => ({
+  getAgentDir: () => '/home/test-user/.pi/agent',
+}));
+
+// ── Module under test ─────────────────────────────────────────────────
+
+import { registerModelDisplay, MODEL_DISPLAY_STATUS_KEY } from './model-display.js';
+
+describe('model-display', () => {
+  /** Tracks event listeners registered by registerModelDisplay. */
+  const registeredListeners: Record<string, Function> = {};
+  /** Tracks status bar calls: { key: value } */
+  const statusCalls: Record<string, string | undefined> = {};
+
+  /** Mock ExtensionAPI with on() and a minimal context with ui. */
+  let mockPi: any;
+  let mockCtx: any;
+
+  beforeEach(() => {
+    // Reset tracking
+    Object.keys(registeredListeners).forEach(k => delete registeredListeners[k]);
+    Object.keys(statusCalls).forEach(k => delete statusCalls[k]);
+
+    // Mock extension context
+    mockCtx = {
+      ui: {
+        setStatus: vi.fn((key: string, value: string | undefined) => {
+          statusCalls[key] = value;
+        }),
+      },
+    };
+
+    // Mock Pi API
+    mockPi = {
+      on: vi.fn((event: string, handler: Function) => {
+        registeredListeners[event] = handler;
+      }),
+    };
+  });
+
+  // ── Exports ───────────────────────────────────────────────────────
+
+  it('exports MODEL_DISPLAY_STATUS_KEY', () => {
+    expect(MODEL_DISPLAY_STATUS_KEY).toBe('worklog-model');
+  });
+
+  it('exports registerModelDisplay as a function', () => {
+    expect(typeof registerModelDisplay).toBe('function');
+  });
+
+  // ── Event registration ───────────────────────────────────────────
+
+  it('registers model_select listener', () => {
+    registerModelDisplay(mockPi);
+    expect(mockPi.on).toHaveBeenCalledWith('model_select', expect.any(Function));
+    expect(registeredListeners['model_select']).toBeDefined();
+  });
+
+  it('registers after_provider_response listener', () => {
+    registerModelDisplay(mockPi);
+    expect(mockPi.on).toHaveBeenCalledWith('after_provider_response', expect.any(Function));
+    expect(registeredListeners['after_provider_response']).toBeDefined();
+  });
+
+  // ── model_select behavior ────────────────────────────────────────
+
+  it('model_select sets status with selected model id and pending provider', () => {
+    registerModelDisplay(mockPi);
+    const handler = registeredListeners['model_select'];
+
+    handler({ model: { id: 'plan', provider: 'openai' } }, mockCtx);
+
+    expect(mockCtx.ui.setStatus).toHaveBeenCalledWith(
+      'worklog-model',
+      'plan → (pending)',
+    );
+  });
+
+  it('model_select without model id does not crash', () => {
+    registerModelDisplay(mockPi);
+    const handler = registeredListeners['model_select'];
+
+    handler({ model: null }, mockCtx);
+
+    // No status call expected because there's no model to display
+    expect(mockCtx.ui.setStatus).not.toHaveBeenCalled();
+  });
+
+  it('model_select updates status when after_provider_response was previously received', () => {
+    registerModelDisplay(mockPi);
+    const modelHandler = registeredListeners['model_select'];
+    const responseHandler = registeredListeners['after_provider_response'];
+
+    // First, receive a provider response
+    responseHandler({ headers: { 'x-resolved-model': 'openai/gpt-4' } }, mockCtx);
+    expect(mockCtx.ui.setStatus).toHaveBeenLastCalledWith(
+      'worklog-model',
+      '→ openai/gpt-4',
+    );
+
+    // Then, select a model — should show model → provider
+    modelHandler({ model: { id: 'plan' } }, mockCtx);
+    expect(mockCtx.ui.setStatus).toHaveBeenLastCalledWith(
+      'worklog-model',
+      'plan → openai/gpt-4',
+    );
+  });
+
+  // ── after_provider_response behavior ──────────────────────────────
+
+  it('after_provider_response extracts X-Resolved-Model from headers', () => {
+    registerModelDisplay(mockPi);
+    const handler = registeredListeners['after_provider_response'];
+
+    handler({ headers: { 'x-resolved-model': 'openai/gpt-4' } }, mockCtx);
+
+    expect(mockCtx.ui.setStatus).toHaveBeenCalledWith(
+      'worklog-model',
+      '→ openai/gpt-4',
+    );
+  });
+
+  it('after_provider_response handles uppercase X-Resolved-Model header', () => {
+    registerModelDisplay(mockPi);
+    const handler = registeredListeners['after_provider_response'];
+
+    handler({ headers: { 'X-Resolved-Model': 'anthropic/claude-3' } }, mockCtx);
+
+    expect(mockCtx.ui.setStatus).toHaveBeenCalledWith(
+      'worklog-model',
+      '→ anthropic/claude-3',
+    );
+  });
+
+  it('after_provider_response without resolved model header does not crash', () => {
+    registerModelDisplay(mockPi);
+    const handler = registeredListeners['after_provider_response'];
+
+    handler({ headers: {} }, mockCtx);
+
+    // No call expected because nothing to update if no context is set
+    expect(mockCtx.ui.setStatus).not.toHaveBeenCalled();
+  });
+
+  it('after_provider_response shows model → provider when model was previously selected', () => {
+    registerModelDisplay(mockPi);
+    const modelHandler = registeredListeners['model_select'];
+    const responseHandler = registeredListeners['after_provider_response'];
+
+    // First, select a model
+    modelHandler({ model: { id: 'code' } }, mockCtx);
+
+    // Then, receive a provider response
+    responseHandler({ headers: { 'x-resolved-model': 'anthropic/claude-sonnet-4' } }, mockCtx);
+
+    expect(mockCtx.ui.setStatus).toHaveBeenLastCalledWith(
+      'worklog-model',
+      'code → anthropic/claude-sonnet-4',
+    );
+  });
+
+  // ── Status key verification ───────────────────────────────────────
+
+  it('uses status key "worklog-model" not "model-display"', () => {
+    registerModelDisplay(mockPi);
+    const handler = registeredListeners['model_select'];
+
+    handler({ model: { id: 'plan', provider: 'openai' } }, mockCtx);
+
+    // Verify the old key is NOT used
+    expect(statusCalls['model-display']).toBeUndefined();
+    // Verify the new key IS used
+    expect(statusCalls['worklog-model']).toBeDefined();
+  });
+
+  // ── Edge cases ───────────────────────────────────────────────────
+
+  it('handles multiple model_select calls, overwriting previous model', () => {
+    registerModelDisplay(mockPi);
+    const handler = registeredListeners['model_select'];
+
+    handler({ model: { id: 'plan' } }, mockCtx);
+    handler({ model: { id: 'code' } }, mockCtx);
+
+    expect(mockCtx.ui.setStatus).toHaveBeenLastCalledWith(
+      'worklog-model',
+      'code → (pending)',
+    );
+  });
+
+  it('handles multiple after_provider_response calls, overwriting previous resolved model', () => {
+    registerModelDisplay(mockPi);
+    const modelHandler = registeredListeners['model_select'];
+    const responseHandler = registeredListeners['after_provider_response'];
+
+    modelHandler({ model: { id: 'code' } }, mockCtx);
+    responseHandler({ headers: { 'x-resolved-model': 'openai/gpt-4' } }, mockCtx);
+    responseHandler({ headers: { 'x-resolved-model': 'anthropic/claude-sonnet-4' } }, mockCtx);
+
+    expect(mockCtx.ui.setStatus).toHaveBeenLastCalledWith(
+      'worklog-model',
+      'code → anthropic/claude-sonnet-4',
+    );
+  });
+
+  it('after_provider_response with selectedModel=null shows → provider only', () => {
+    registerModelDisplay(mockPi);
+    const handler = registeredListeners['after_provider_response'];
+
+    // Without any prior model_select
+    handler({ headers: { 'x-resolved-model': 'openai/gpt-4' } }, mockCtx);
+
+    expect(mockCtx.ui.setStatus).toHaveBeenCalledWith(
+      'worklog-model',
+      '→ openai/gpt-4',
+    );
+  });
+});
diff --git a/packages/tui/extensions/model-display.ts b/packages/tui/extensions/Worklog/model-display.ts
similarity index 56%
rename from packages/tui/extensions/model-display.ts
rename to packages/tui/extensions/Worklog/model-display.ts
index 4b3a61f8..6e00ccf9 100644
--- a/packages/tui/extensions/model-display.ts
+++ b/packages/tui/extensions/Worklog/model-display.ts
@@ -1,8 +1,9 @@
 /**
- * Model Display Extension (LP-0MR4ZIGDT004A3E1)
+ * Model display module for the Worklog Pi extension.
  *
  * Displays the resolved provider/model from llama-proxy responses in Pi's
- * status bar.
+ * status bar. Integrated from the standalone model-display extension
+ * (previously at packages/tui/extensions/model-display.ts).
  *
  * Shows the format: `<selected-model> → <provider>/<model-id>`
  *
@@ -15,9 +16,27 @@
  * Requires the llama-proxy to be configured as Pi's provider and to include
  * the `X-Resolved-Model` header in its responses (see LP-0MR4ZIGDT004A3E1).
  */
-import type { ExtensionAPI } from "@earendil-works/pi-coding-agent";
 
-export default function (pi: ExtensionAPI) {
+import type { ExtensionAPI } from '@earendil-works/pi-coding-agent';
+
+/**
+ * Status key used for the model/provider display in the footer.
+ * Changed from "model-display" when this module was integrated into the
+ * Worklog extension, to avoid conflicts with other status entries.
+ */
+export const MODEL_DISPLAY_STATUS_KEY = 'worklog-model';
+
+/**
+ * Register model-display event handlers with a Pi extension instance.
+ *
+ * Sets up listeners for:
+ * - `model_select` — tracks the currently selected Pi model alias
+ * - `after_provider_response` — reads the `X-Resolved-Model` header from
+ *   llama-proxy responses and updates the status bar
+ *
+ * @param pi - The ExtensionAPI instance
+ */
+export function registerModelDisplay(pi: ExtensionAPI): void {
   // Track the currently selected model alias (e.g., "plan", "code")
   let selectedModel: string | null = null;
 
@@ -27,16 +46,16 @@ export default function (pi: ExtensionAPI) {
   // Update the status display
   function updateStatus(ctx: { ui: { setStatus: (key: string, value: string) => void } }) {
     if (selectedModel && resolvedModel) {
-      ctx.ui.setStatus("model-display", `${selectedModel} → ${resolvedModel}`);
+      ctx.ui.setStatus(MODEL_DISPLAY_STATUS_KEY, `${selectedModel} → ${resolvedModel}`);
     } else if (selectedModel) {
-      ctx.ui.setStatus("model-display", `${selectedModel} → (pending)`);
+      ctx.ui.setStatus(MODEL_DISPLAY_STATUS_KEY, `${selectedModel} → (pending)`);
     } else if (resolvedModel) {
-      ctx.ui.setStatus("model-display", `→ ${resolvedModel}`);
+      ctx.ui.setStatus(MODEL_DISPLAY_STATUS_KEY, `→ ${resolvedModel}`);
     }
   }
 
   // Track model selection changes
-  pi.on("model_select", (_event, ctx) => {
+  pi.on('model_select', (_event, ctx) => {
     // event.model has provider and id fields
     const model = _event.model;
     if (model && model.id) {
@@ -46,12 +65,12 @@ export default function (pi: ExtensionAPI) {
   });
 
   // Extract resolved model from proxy response headers
-  pi.on("after_provider_response", (event, ctx) => {
+  pi.on('after_provider_response', (event, ctx) => {
     // event.headers contains the normalized response headers
     // X-Resolved-Model header is lowercase in normalized form
     const headers = event.headers as Record<string, string>;
     const headerValue: string | undefined =
-      headers["x-resolved-model"] ?? headers["X-Resolved-Model"];
+      headers['x-resolved-model'] ?? headers['X-Resolved-Model'];
 
     if (headerValue) {
       resolvedModel = headerValue;
diff --git a/packages/tui/extensions/package.json b/packages/tui/extensions/package.json
index fc32cdb4..dded3f5d 100644
--- a/packages/tui/extensions/package.json
+++ b/packages/tui/extensions/package.json
@@ -6,8 +6,7 @@
   "description": "Worklog Pi TUI extension — Work item browser for Pi coding agent",
   "pi": {
     "extensions": [
-      "./Worklog/index.ts",
-      "./model-display.ts"
+      "./Worklog/index.ts"
     ]
   }
 }

From 2a34e714f206642c44d41d6cf82e78db858cec5a Mon Sep 17 00:00:00 2001
From: Sorra the Orc <>
Date: Wed, 8 Jul 2026 17:11:57 +0100
Subject: [PATCH 3/8] WL-0MRC91VIN0089UXD: Show only provider/model in status
 bar, remove model alias prefix
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Changes the Worklog extension model status bar display from
'<selected-model> → <provider>/<model-id>' to just '<provider>/<model-id>'.

- Removed model alias display (e.g., 'plan' prefix) from status bar
- Changed (pending) fallback from '<model> → (pending)' to '(pending)'
- Removed '→' arrow prefix from provider-only display
- model_select listener now only tracks model presence for (pending) state
- Updated all 15 tests to reflect new display format
- No changes to X-Resolved-Model header parsing or event listeners
---
 .../extensions/Worklog/model-display.test.ts  | 44 +++++++++----------
 .../tui/extensions/Worklog/model-display.ts   | 25 ++++++-----
 2 files changed, 37 insertions(+), 32 deletions(-)

diff --git a/packages/tui/extensions/Worklog/model-display.test.ts b/packages/tui/extensions/Worklog/model-display.test.ts
index a502a174..b03e0183 100644
--- a/packages/tui/extensions/Worklog/model-display.test.ts
+++ b/packages/tui/extensions/Worklog/model-display.test.ts
@@ -4,11 +4,11 @@
  * Verifies that:
  * 1. registerModelDisplay() sets up event listeners for model_select and
  *    after_provider_response.
- * 2. model_select updates the status bar with the selected model alias.
+ * 2. The status display shows only the resolved provider/model (no model alias).
  * 3. after_provider_response updates the status bar with the resolved
  *    provider/model from X-Resolved-Model header.
- * 4. The status bar key is "worklog-model" (not "model-display").
- * 5. Fallback display states work correctly (pending model, no model).
+ * 4. When no proxy response has been received yet, the status bar shows (pending).
+ * 5. The status bar key is "worklog-model" (not "model-display").
  * 6. The module exports the MODEL_DISPLAY_STATUS_KEY constant.
  *
  * Run: npx vitest run packages/tui/extensions/Worklog/model-display.test.ts
@@ -84,7 +84,7 @@ describe('model-display', () => {
 
   // ── model_select behavior ────────────────────────────────────────
 
-  it('model_select sets status with selected model id and pending provider', () => {
+  it('model_select sets status to (pending) when no response yet', () => {
     registerModelDisplay(mockPi);
     const handler = registeredListeners['model_select'];
 
@@ -92,21 +92,21 @@ describe('model-display', () => {
 
     expect(mockCtx.ui.setStatus).toHaveBeenCalledWith(
       'worklog-model',
-      'plan → (pending)',
+      '(pending)',
     );
   });
 
-  it('model_select without model id does not crash', () => {
+  it('model_select without model id does not crash and does not set status', () => {
     registerModelDisplay(mockPi);
     const handler = registeredListeners['model_select'];
 
     handler({ model: null }, mockCtx);
 
-    // No status call expected because there's no model to display
+    // No status call expected because there's nothing to display
     expect(mockCtx.ui.setStatus).not.toHaveBeenCalled();
   });
 
-  it('model_select updates status when after_provider_response was previously received', () => {
+  it('model_select no longer shows model alias when after_provider_response was previously received', () => {
     registerModelDisplay(mockPi);
     const modelHandler = registeredListeners['model_select'];
     const responseHandler = registeredListeners['after_provider_response'];
@@ -115,20 +115,20 @@ describe('model-display', () => {
     responseHandler({ headers: { 'x-resolved-model': 'openai/gpt-4' } }, mockCtx);
     expect(mockCtx.ui.setStatus).toHaveBeenLastCalledWith(
       'worklog-model',
-      '→ openai/gpt-4',
+      'openai/gpt-4',
     );
 
-    // Then, select a model — should show model → provider
+    // Then, select a model — should still show provider/model only (no model alias)
     modelHandler({ model: { id: 'plan' } }, mockCtx);
     expect(mockCtx.ui.setStatus).toHaveBeenLastCalledWith(
       'worklog-model',
-      'plan → openai/gpt-4',
+      'openai/gpt-4',
     );
   });
 
   // ── after_provider_response behavior ──────────────────────────────
 
-  it('after_provider_response extracts X-Resolved-Model from headers', () => {
+  it('after_provider_response extracts X-Resolved-Model from headers (no arrow prefix)', () => {
     registerModelDisplay(mockPi);
     const handler = registeredListeners['after_provider_response'];
 
@@ -136,7 +136,7 @@ describe('model-display', () => {
 
     expect(mockCtx.ui.setStatus).toHaveBeenCalledWith(
       'worklog-model',
-      '→ openai/gpt-4',
+      'openai/gpt-4',
     );
   });
 
@@ -148,7 +148,7 @@ describe('model-display', () => {
 
     expect(mockCtx.ui.setStatus).toHaveBeenCalledWith(
       'worklog-model',
-      '→ anthropic/claude-3',
+      'anthropic/claude-3',
     );
   });
 
@@ -158,11 +158,11 @@ describe('model-display', () => {
 
     handler({ headers: {} }, mockCtx);
 
-    // No call expected because nothing to update if no context is set
+    // No call expected because nothing to update
     expect(mockCtx.ui.setStatus).not.toHaveBeenCalled();
   });
 
-  it('after_provider_response shows model → provider when model was previously selected', () => {
+  it('after_provider_response shows provider/model (no model alias) when model was previously selected', () => {
     registerModelDisplay(mockPi);
     const modelHandler = registeredListeners['model_select'];
     const responseHandler = registeredListeners['after_provider_response'];
@@ -175,7 +175,7 @@ describe('model-display', () => {
 
     expect(mockCtx.ui.setStatus).toHaveBeenLastCalledWith(
       'worklog-model',
-      'code → anthropic/claude-sonnet-4',
+      'anthropic/claude-sonnet-4',
     );
   });
 
@@ -195,7 +195,7 @@ describe('model-display', () => {
 
   // ── Edge cases ───────────────────────────────────────────────────
 
-  it('handles multiple model_select calls, overwriting previous model', () => {
+  it('handles multiple model_select calls, staying in pending state', () => {
     registerModelDisplay(mockPi);
     const handler = registeredListeners['model_select'];
 
@@ -204,7 +204,7 @@ describe('model-display', () => {
 
     expect(mockCtx.ui.setStatus).toHaveBeenLastCalledWith(
       'worklog-model',
-      'code → (pending)',
+      '(pending)',
     );
   });
 
@@ -219,11 +219,11 @@ describe('model-display', () => {
 
     expect(mockCtx.ui.setStatus).toHaveBeenLastCalledWith(
       'worklog-model',
-      'code → anthropic/claude-sonnet-4',
+      'anthropic/claude-sonnet-4',
     );
   });
 
-  it('after_provider_response with selectedModel=null shows → provider only', () => {
+  it('after_provider_response with no prior model_select shows provider/model only', () => {
     registerModelDisplay(mockPi);
     const handler = registeredListeners['after_provider_response'];
 
@@ -232,7 +232,7 @@ describe('model-display', () => {
 
     expect(mockCtx.ui.setStatus).toHaveBeenCalledWith(
       'worklog-model',
-      '→ openai/gpt-4',
+      'openai/gpt-4',
     );
   });
 });
diff --git a/packages/tui/extensions/Worklog/model-display.ts b/packages/tui/extensions/Worklog/model-display.ts
index 6e00ccf9..8aeac7c5 100644
--- a/packages/tui/extensions/Worklog/model-display.ts
+++ b/packages/tui/extensions/Worklog/model-display.ts
@@ -5,11 +5,13 @@
  * status bar. Integrated from the standalone model-display extension
  * (previously at packages/tui/extensions/model-display.ts).
  *
- * Shows the format: `<selected-model> → <provider>/<model-id>`
+ * Shows the format: `<provider>/<model-id>` (e.g., `openai/gpt-4`)
+ * without the Pi model alias prefix.
  *
  * - Listens to `after_provider_response` to read the `X-Resolved-Model` header
  *   that the llama-proxy adds to every proxy-routed request.
- * - Listens to `model_select` to track the currently selected Pi model alias.
+ * - Listens to `model_select` to track whether a model is selected (for the
+ *   `(pending)` fallback state), but does not display the model alias.
  * - Updates the status bar on every request so the user always sees which
  *   provider and model actually served their request.
  *
@@ -30,14 +32,18 @@ export const MODEL_DISPLAY_STATUS_KEY = 'worklog-model';
  * Register model-display event handlers with a Pi extension instance.
  *
  * Sets up listeners for:
- * - `model_select` — tracks the currently selected Pi model alias
+ * - `model_select` — tracks whether a model is selected (for `(pending)` state)
  * - `after_provider_response` — reads the `X-Resolved-Model` header from
  *   llama-proxy responses and updates the status bar
  *
+ * The status display shows only the resolved provider/model from the
+ * `X-Resolved-Model` header. When no response has been received yet but a
+ * model is selected, it shows `(pending)`. The Pi model alias is not shown.
+ *
  * @param pi - The ExtensionAPI instance
  */
 export function registerModelDisplay(pi: ExtensionAPI): void {
-  // Track the currently selected model alias (e.g., "plan", "code")
+  // Track whether a model is selected (used for the (pending) fallback state)
   let selectedModel: string | null = null;
 
   // Track the resolved model from the last proxy response
@@ -45,16 +51,15 @@ export function registerModelDisplay(pi: ExtensionAPI): void {
 
   // Update the status display
   function updateStatus(ctx: { ui: { setStatus: (key: string, value: string) => void } }) {
-    if (selectedModel && resolvedModel) {
-      ctx.ui.setStatus(MODEL_DISPLAY_STATUS_KEY, `${selectedModel} → ${resolvedModel}`);
+    if (resolvedModel) {
+      ctx.ui.setStatus(MODEL_DISPLAY_STATUS_KEY, resolvedModel);
     } else if (selectedModel) {
-      ctx.ui.setStatus(MODEL_DISPLAY_STATUS_KEY, `${selectedModel} → (pending)`);
-    } else if (resolvedModel) {
-      ctx.ui.setStatus(MODEL_DISPLAY_STATUS_KEY, `→ ${resolvedModel}`);
+      ctx.ui.setStatus(MODEL_DISPLAY_STATUS_KEY, '(pending)');
     }
+    // If neither resolvedModel nor selectedModel, no status call (nothing to display)
   }
 
-  // Track model selection changes
+  // Track model selection changes (used only for (pending) fallback)
   pi.on('model_select', (_event, ctx) => {
     // event.model has provider and id fields
     const model = _event.model;

From f0fad16410901724fe4508d15cebaf2391f22e87 Mon Sep 17 00:00:00 2001
From: Sorra the Orc <>
Date: Wed, 8 Jul 2026 23:30:45 +0100
Subject: [PATCH 4/8] Add audit approve/reject shortcuts and update docs

- Add a+y (audit approve) and a+r (audit reject) chord shortcuts
- Rename a single-key audit to a,a chord (audit automatic)
- Add shortcuts.json source path note to README.md
- Document --ready-to-close yes/no usage in AUDIT_STATUS.md
- Bump version to 1.0.1
- Update tests for new entry/chord counts and removed a key
---
 docs/AUDIT_STATUS.md                          |  9 +++++++
 packages/tui/extensions/README.md             |  3 +++
 .../Worklog/shortcut-config.test.ts           | 25 ++++++-------------
 .../tui/extensions/Worklog/shortcuts.json     | 20 +++++++++++++--
 src/version.ts                                |  2 +-
 tests/cli/status.test.ts                      |  2 +-
 6 files changed, 40 insertions(+), 21 deletions(-)

diff --git a/docs/AUDIT_STATUS.md b/docs/AUDIT_STATUS.md
index 591cd694..ebd5eb3f 100644
--- a/docs/AUDIT_STATUS.md
+++ b/docs/AUDIT_STATUS.md
@@ -34,8 +34,17 @@ All acceptance criteria verified."
 
 # Set audit via the new dedicated command
 wl audit-set SA-123 --ready-to-close --summary "All acceptance criteria verified." --author agent
+
+# Mark an audit as approved by manual review
+wl audit-set SA-123 --ready-to-close yes --summary "Manual review approved." --author "reviewer-name"
+
+# Mark an audit as not yet ready (e.g., after a failed review)
+wl audit-set SA-123 --ready-to-close no --summary "Manual review failed: criteria X not met."
 ```
 
+> **Note:** `--ready-to-close` accepts both a bare flag (`--ready-to-close`) which defaults to `yes`,
+> and an explicit value (`--ready-to-close yes` or `--ready-to-close no`).
+
 ### Viewing audit results
 
 ```bash
diff --git a/packages/tui/extensions/README.md b/packages/tui/extensions/README.md
index 6732ed9a..7011ac60 100644
--- a/packages/tui/extensions/README.md
+++ b/packages/tui/extensions/README.md
@@ -382,6 +382,9 @@ The `/wl` command registers `getArgumentCompletions`, so Pi's editor shows autoc
 
 ## Shortcuts
 
+Source config file: `packages/tui/extensions/Worklog/shortcuts.json`
+(installed to `~/.pi/agent/extensions/worklog/Worklog/shortcuts.json`).
+
 The `shortcuts.json` config file defines a **config-driven shortcut system** that allows keyboard shortcuts in the Pi extension's worklog browse views (list and detail) to be expressed declaratively rather than hardcoded.
 
 ### Schema
diff --git a/packages/tui/extensions/Worklog/shortcut-config.test.ts b/packages/tui/extensions/Worklog/shortcut-config.test.ts
index 5b27c7f9..a6727cef 100644
--- a/packages/tui/extensions/Worklog/shortcut-config.test.ts
+++ b/packages/tui/extensions/Worklog/shortcut-config.test.ts
@@ -210,7 +210,7 @@ describe('loadShortcutConfig', () => {
   it('loads valid entries from shortcuts.json', () => {
     const registry = loadShortcutConfig();
     const entries = registry.getEntries();
-    expect(entries).toHaveLength(15);
+    expect(entries).toHaveLength(17);
 
     const createEntry = entries.find(e => e.key === 'c');
     expect(createEntry).toBeDefined();
@@ -244,15 +244,7 @@ describe('loadShortcutConfig', () => {
     expect(intakeEntry!.label).toBe('intake');
     expect(intakeEntry!.description).toBe('Ensure that the selected item is reasonably well defined in terms of objectives.');
 
-    const auditEntry = entries.find(e => e.key === 'a');
-    expect(auditEntry).toBeDefined();
-    expect(auditEntry!.command).toBe('/skill:audit <id>');
-    expect(auditEntry!.view).toBe('both');
-    expect(auditEntry!.stages).toEqual(['in_progress', 'in_review']);
-    expect(auditEntry!.label).toBe('audit');
-    expect(auditEntry!.description).toBe('Run an audit on the selected work item');
-
-    expect(entries.filter(e => e.key === '').length).toBe(9); // 9 chord entries have empty key
+    expect(entries.filter(e => e.key === '').length).toBe(12); // 12 chord entries have empty key
   });
 
   it('has no duplicate key+view or chord+view combinations in shortcuts.json', () => {
@@ -297,18 +289,17 @@ describe('loadShortcutConfig', () => {
     expect(registry.lookup('n', 'detail', 'idea')).toBe('/intake <id>');
     expect(registry.lookup('n', 'detail', 'intake_complete')).toBeUndefined();
 
-    // 'a' (audit) has stages: ['in_progress', 'in_review']
-    expect(registry.lookup('a', 'list', 'in_review')).toBe('/skill:audit <id>');
-    expect(registry.lookup('a', 'detail', 'in_review')).toBe('/skill:audit <id>');
-    expect(registry.lookup('a', 'list', 'in_progress')).toBe('/skill:audit <id>');
-    expect(registry.lookup('a', 'list', 'idea')).toBeUndefined();
+    // 'a' is now a chord leader (a,a for audit, a+y for approve, a+r for reject),
+    // so single-key lookup for 'a' returns undefined
+    expect(registry.lookup('a', 'list', 'in_review')).toBeUndefined();
+    expect(registry.lookup('a', 'detail', 'in_review')).toBeUndefined();
+    expect(registry.lookup('a', 'list', 'in_progress')).toBeUndefined();
 
     // Without stage parameter, entries with stages constraint still work
     // (backward compatible when calling without stage)
     expect(registry.lookup('n', 'detail')).toBe('/intake <id>');
     expect(registry.lookup('i', 'list')).toBe('/skill:implement <id>');
     expect(registry.lookup('p', 'list')).toBe('/plan <id>');
-    expect(registry.lookup('a', 'list')).toBe('/skill:audit <id>');
   });
 
   it('loads chord entries from shortcuts.json', () => {
@@ -316,7 +307,7 @@ describe('loadShortcutConfig', () => {
     const entries = registry.getEntries();
 
     const upChords = registry.getChordEntries();
-    expect(upChords).toHaveLength(9);
+    expect(upChords).toHaveLength(12);
 
     const upEntry = upChords.find((e: any) =>
       Array.isArray((e as any).chord) && (e as any).chord[0] === 'u' && (e as any).chord[1] === 'p',
diff --git a/packages/tui/extensions/Worklog/shortcuts.json b/packages/tui/extensions/Worklog/shortcuts.json
index e6ad9611..f36582f2 100644
--- a/packages/tui/extensions/Worklog/shortcuts.json
+++ b/packages/tui/extensions/Worklog/shortcuts.json
@@ -29,13 +29,29 @@
     "stages": ["intake_complete", "plan_complete", "in_progress"],
     "label": "implement",
     "description": "Run the implement workflow on the selected work item"
+  },
+    {
+    "chord": ["a", "y"],
+    "command": "!!wl audit-set <id> --ready-to-close yes --summary 'Approved by manual review'",
+    "view": "both",
+    "stages": ["in_review"],
+    "label": "audit approve",
+    "description": "Approve the in_review item by manual audit."
+  },
+  {
+    "chord": ["a", "r"],
+    "command": "!!wl audit-set <id> --ready-to-close no --summary 'Rejected by manual review. <reason>'",
+    "view": "both",
+    "stages": ["in_review"],
+    "label": "audit reject",
+    "description": "Reject the in_review item by manual audit."
   },
   {
-    "key": "a",
+    "chord": ["a", "a"],
     "command": "/skill:audit <id>",
     "view": "both",
     "stages": ["in_progress", "in_review"],
-    "label": "audit",
+    "label": "audit automatic",
     "description": "Run an audit on the selected work item"
   },
   {
diff --git a/src/version.ts b/src/version.ts
index 14260576..50fff331 100644
--- a/src/version.ts
+++ b/src/version.ts
@@ -1,2 +1,2 @@
 // Auto-generated; do not edit.
-export const WORKLOG_VERSION = '1.0.0';
+export const WORKLOG_VERSION = '1.0.1';
diff --git a/tests/cli/status.test.ts b/tests/cli/status.test.ts
index cd5b60da..5d0abe00 100644
--- a/tests/cli/status.test.ts
+++ b/tests/cli/status.test.ts
@@ -108,7 +108,7 @@ describe('CLI Status Tests', () => {
 
     expect(stdout).toContain('Worklog System Status');
     expect(stdout).toContain('Initialized: Yes');
-    expect(stdout).toContain('Version: 1.0.0');
+    expect(stdout).toContain('Version: 1.0.1');
     expect(stdout).toContain('Configuration:');
     expect(stdout).toContain('Database Summary:');
     expect(stdout).toContain('Work Items:');

From c2848861f240f64928883a5a4b45cdd051230764 Mon Sep 17 00:00:00 2001
From: Sorra the Orc <>
Date: Wed, 8 Jul 2026 23:48:51 +0100
Subject: [PATCH 5/8] WL-0MRC91VIN0089UXD: Ensure provider/model sorts first in
 extension status line by renaming key to worklog-0model

Pi renders extension statuses sorted alphabetically by key. The previous key
'worklog-model' sorted last (after worklog-activity and worklog-auto-inject),
leaving the provider/model at the end of the extension status line.

Renamed to 'worklog-0model' so '0' < 'a' in localeCompare, placing the
provider/model at the start of the line.

Also updated all 15 test assertions and the export constant.
---
 .../extensions/Worklog/model-display.test.ts  | 27 ++++++++++---------
 .../tui/extensions/Worklog/model-display.ts   |  4 ++-
 2 files changed, 17 insertions(+), 14 deletions(-)

diff --git a/packages/tui/extensions/Worklog/model-display.test.ts b/packages/tui/extensions/Worklog/model-display.test.ts
index b03e0183..b76655b4 100644
--- a/packages/tui/extensions/Worklog/model-display.test.ts
+++ b/packages/tui/extensions/Worklog/model-display.test.ts
@@ -8,7 +8,8 @@
  * 3. after_provider_response updates the status bar with the resolved
  *    provider/model from X-Resolved-Model header.
  * 4. When no proxy response has been received yet, the status bar shows (pending).
- * 5. The status bar key is "worklog-model" (not "model-display").
+ * 5. The status bar key is "worklog-0model" (not "model-display"),
+ *    with the `0` prefix ensuring it sorts first in the extension status line.
  * 6. The module exports the MODEL_DISPLAY_STATUS_KEY constant.
  *
  * Run: npx vitest run packages/tui/extensions/Worklog/model-display.test.ts
@@ -61,7 +62,7 @@ describe('model-display', () => {
   // ── Exports ───────────────────────────────────────────────────────
 
   it('exports MODEL_DISPLAY_STATUS_KEY', () => {
-    expect(MODEL_DISPLAY_STATUS_KEY).toBe('worklog-model');
+    expect(MODEL_DISPLAY_STATUS_KEY).toBe('worklog-0model');
   });
 
   it('exports registerModelDisplay as a function', () => {
@@ -91,7 +92,7 @@ describe('model-display', () => {
     handler({ model: { id: 'plan', provider: 'openai' } }, mockCtx);
 
     expect(mockCtx.ui.setStatus).toHaveBeenCalledWith(
-      'worklog-model',
+      'worklog-0model',
       '(pending)',
     );
   });
@@ -114,14 +115,14 @@ describe('model-display', () => {
     // First, receive a provider response
     responseHandler({ headers: { 'x-resolved-model': 'openai/gpt-4' } }, mockCtx);
     expect(mockCtx.ui.setStatus).toHaveBeenLastCalledWith(
-      'worklog-model',
+      'worklog-0model',
       'openai/gpt-4',
     );
 
     // Then, select a model — should still show provider/model only (no model alias)
     modelHandler({ model: { id: 'plan' } }, mockCtx);
     expect(mockCtx.ui.setStatus).toHaveBeenLastCalledWith(
-      'worklog-model',
+      'worklog-0model',
       'openai/gpt-4',
     );
   });
@@ -135,7 +136,7 @@ describe('model-display', () => {
     handler({ headers: { 'x-resolved-model': 'openai/gpt-4' } }, mockCtx);
 
     expect(mockCtx.ui.setStatus).toHaveBeenCalledWith(
-      'worklog-model',
+      'worklog-0model',
       'openai/gpt-4',
     );
   });
@@ -147,7 +148,7 @@ describe('model-display', () => {
     handler({ headers: { 'X-Resolved-Model': 'anthropic/claude-3' } }, mockCtx);
 
     expect(mockCtx.ui.setStatus).toHaveBeenCalledWith(
-      'worklog-model',
+      'worklog-0model',
       'anthropic/claude-3',
     );
   });
@@ -174,14 +175,14 @@ describe('model-display', () => {
     responseHandler({ headers: { 'x-resolved-model': 'anthropic/claude-sonnet-4' } }, mockCtx);
 
     expect(mockCtx.ui.setStatus).toHaveBeenLastCalledWith(
-      'worklog-model',
+      'worklog-0model',
       'anthropic/claude-sonnet-4',
     );
   });
 
   // ── Status key verification ───────────────────────────────────────
 
-  it('uses status key "worklog-model" not "model-display"', () => {
+  it('uses status key "worklog-0model" (sorts first) not "model-display"', () => {
     registerModelDisplay(mockPi);
     const handler = registeredListeners['model_select'];
 
@@ -190,7 +191,7 @@ describe('model-display', () => {
     // Verify the old key is NOT used
     expect(statusCalls['model-display']).toBeUndefined();
     // Verify the new key IS used
-    expect(statusCalls['worklog-model']).toBeDefined();
+    expect(statusCalls['worklog-0model']).toBeDefined();
   });
 
   // ── Edge cases ───────────────────────────────────────────────────
@@ -203,7 +204,7 @@ describe('model-display', () => {
     handler({ model: { id: 'code' } }, mockCtx);
 
     expect(mockCtx.ui.setStatus).toHaveBeenLastCalledWith(
-      'worklog-model',
+      'worklog-0model',
       '(pending)',
     );
   });
@@ -218,7 +219,7 @@ describe('model-display', () => {
     responseHandler({ headers: { 'x-resolved-model': 'anthropic/claude-sonnet-4' } }, mockCtx);
 
     expect(mockCtx.ui.setStatus).toHaveBeenLastCalledWith(
-      'worklog-model',
+      'worklog-0model',
       'anthropic/claude-sonnet-4',
     );
   });
@@ -231,7 +232,7 @@ describe('model-display', () => {
     handler({ headers: { 'x-resolved-model': 'openai/gpt-4' } }, mockCtx);
 
     expect(mockCtx.ui.setStatus).toHaveBeenCalledWith(
-      'worklog-model',
+      'worklog-0model',
       'openai/gpt-4',
     );
   });
diff --git a/packages/tui/extensions/Worklog/model-display.ts b/packages/tui/extensions/Worklog/model-display.ts
index 8aeac7c5..31a0e2b0 100644
--- a/packages/tui/extensions/Worklog/model-display.ts
+++ b/packages/tui/extensions/Worklog/model-display.ts
@@ -23,10 +23,12 @@ import type { ExtensionAPI } from '@earendil-works/pi-coding-agent';
 
 /**
  * Status key used for the model/provider display in the footer.
+ * The `0` prefix ensures this status sorts first in the extension status
+ * line (localeCompare order), placing the provider/model at the start.
  * Changed from "model-display" when this module was integrated into the
  * Worklog extension, to avoid conflicts with other status entries.
  */
-export const MODEL_DISPLAY_STATUS_KEY = 'worklog-model';
+export const MODEL_DISPLAY_STATUS_KEY = 'worklog-0model';
 
 /**
  * Register model-display event handlers with a Pi extension instance.

From e403b257db9a778ad4dbba0f4e6b37976a139677 Mon Sep 17 00:00:00 2001
From: Sorra the Orc <>
Date: Wed, 8 Jul 2026 23:51:15 +0100
Subject: [PATCH 6/8] =?UTF-8?q?WL-0MRCN2HF5003HSG9:=20Change=20audit=20que?=
 =?UTF-8?q?stion=20mark=20icon=20from=20red=20(=E2=9D=93)=20to=20grey=20(?=
 =?UTF-8?q?=E2=9D=94)?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Changes:
- src/icons.ts: Changed AUDIT_ICON.unknown from \u{2753} (❓, red question
  mark) to \u{2754} (❔, white/grey question mark ornament)
- tests/unit/icons.test.ts: Updated 2 auditIcon assertions to expect \u{2754}
- packages/tui/extensions/Worklog/terminal-utils.test.ts: Updated visible-width
  test to use the new character
---
 packages/tui/extensions/Worklog/terminal-utils.test.ts | 2 +-
 src/icons.ts                                           | 2 +-
 tests/unit/icons.test.ts                               | 4 ++--
 3 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/packages/tui/extensions/Worklog/terminal-utils.test.ts b/packages/tui/extensions/Worklog/terminal-utils.test.ts
index 35702d3a..269fb38b 100644
--- a/packages/tui/extensions/Worklog/terminal-utils.test.ts
+++ b/packages/tui/extensions/Worklog/terminal-utils.test.ts
@@ -170,7 +170,7 @@ describe('terminal-utils', () => {
 
     it('handles mixed icons with VS16 sequences correctly', () => {
       // 🔓 (2) + space (1) + 🛠️ (2) + space (1) + ❓ (2) = 8
-      expect(visibleWidth('\u{1F513} \u{1F6E0}\uFE0F \u{2753}')).toBe(8);
+      expect(visibleWidth('\u{1F513} \u{1F6E0}\uFE0F \u{2754}')).toBe(8);
     });
 
     it('treats ZWJ sequences correctly counting only visible characters', () => {
diff --git a/src/icons.ts b/src/icons.ts
index 010dcdc2..6edc557a 100644
--- a/src/icons.ts
+++ b/src/icons.ts
@@ -287,7 +287,7 @@ function auditKey(result: boolean | null | undefined): string {
 const AUDIT_ICON: Record<string, string> = {
   yes:     '\u{2705}',  // ✅
   no:      '\u{274C}',  // ❌
-  unknown: '\u{2753}',  // ❓
+  unknown: '\u{2754}',  // ❔
 };
 
 const AUDIT_FALLBACK: Record<string, string> = {
diff --git a/tests/unit/icons.test.ts b/tests/unit/icons.test.ts
index ceefa52a..ca42bcc9 100644
--- a/tests/unit/icons.test.ts
+++ b/tests/unit/icons.test.ts
@@ -443,11 +443,11 @@ describe('auditIcon', () => {
   });
 
   it('returns emoji for unknown (null)', () => {
-    expect(auditIcon(null)).toBe('\u{2753}'); // ❓
+    expect(auditIcon(null)).toBe('\u{2754}'); // ❔
   });
 
   it('returns emoji for unknown (undefined)', () => {
-    expect(auditIcon(undefined)).toBe('\u{2753}'); // ❓
+    expect(auditIcon(undefined)).toBe('\u{2754}'); // ❔
   });
 
   describe('with noIcons option', () => {

From f7e750ce4bf331c5485f745ef74b50d18e5c083a Mon Sep 17 00:00:00 2001
From: Sorra the Orc <>
Date: Wed, 8 Jul 2026 23:55:30 +0100
Subject: [PATCH 7/8] Bump version to v1.0.2

---
 package.json | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/package.json b/package.json
index d9941ac9..81db6969 100644
--- a/package.json
+++ b/package.json
@@ -1,6 +1,6 @@
 {
   "name": "worklog",
-  "version": "1.0.1",
+  "version": "1.0.2",
   "description": "A simple experimental issue tracker for AI agents",
   "main": "dist/index.js",
   "type": "module",

From bf8a1c9bfbbfd197523930eca4b956ade1f224d4 Mon Sep 17 00:00:00 2001
From: Sorra the Orc <>
Date: Wed, 8 Jul 2026 23:55:34 +0100
Subject: [PATCH 8/8] Update CHANGELOG.md for v1.0.2

---
 CHANGELOG.md | 9 +++++++++
 1 file changed, 9 insertions(+)
 create mode 100644 CHANGELOG.md

diff --git a/CHANGELOG.md b/CHANGELOG.md
new file mode 100644
index 00000000..ecb771f6
--- /dev/null
+++ b/CHANGELOG.md
@@ -0,0 +1,9 @@
+# Changelog
+
+## v1.0.2 (2026-07-08)
+### Other
+- Integrate model-display extension into the Worklog extension (WL-0MRC3F30E008XJN3)
+- Move provider/model display to start of Worklog extension status line (WL-0MRC91VIN0089UXD)
+- Change audit/review question mark icon from red to grey in selection list (WL-0MRCN2HF5003HSG9)
+- Model display extension is not showing the provider (WL-0MRC3NOVZ000P0P5)
+