test(authority): refresh todo content revision after edit

[mikamikasuki]

Goal And Delivered Outcome

• Outcome basis: A reproducible failure in the canonical Todo authority test on main.
• Goal/source and gap: After a Todo text edit, the test compared the updated record with its original snapshot and required the original content_revision to remain unchanged.
• Observable before to after: On base 3ed5d6b, the focused test fails because the updated text produces digest sha256:3b61f6402feb68632608a56817e5dd4ffdf0b041d60b0d1d40a1d9f0a4305036 while the assertion expects the old digest. The test now derives the expected revision from the edited text and verifies that it differs from the prior revision.
• Issue/task and intended base: No separate issue is required for this self-contained test correction. Target main at 3ed5d6b.

Author Declaration

• Written by: model_agent — OpenAI GPT-6 Luna Medium

Implemented against

• Specification and revision: loopx/control_plane/todos/summary_item.py::todo_text_content_revision at 3ed5d6b; Todo freshness behavior is also described in PR #5911.
• Criteria:

Criterion	Disposition	Symbol / path	Test or command
An edited Todo's content revision is derived from its normalized text and differs from the prior revision	implemented	tests/control_plane/test_local_coordination_authority.py	test_real_canonical_provider_preserves_complete_complex_todo_semantics

• Self-check before submission: Reproduced the stale assertion on the exact base, confirmed the digest function uses normalized Todo text, searched for duplicate work, and reviewed the one-file test-only diff.

Scope And Continuation

• Completed scope and remaining work: Corrected the expected content revision after the test edits provider-owned Todo text; production behavior is unchanged.
• Slice boundary / successor: Complete within this scope.

Validation

• Tested revision: 3b48964
• Run state: finished
• Input classes: synthetic

Check kind	Result	Public-safe evidence
regression_parity	passed	Exact-base focused test failed on the stale digest expectation; the corrected assertion passes on this head.
unit	passed	Focused canonical-provider Todo update test: 1 passed.
static	passed	Ruff and git diff --check passed for the changed test.

• Coverage and gaps: The focused case checks the canonical provider readback after a Todo text update and preserves existing sibling-field assertions. No production code changed.

Frontend / Visual Evidence

• UI impact: none
• Before: N/A
• After: N/A
• States and viewports shown: N/A
• Source data: synthetic
• Attention review: N/A

Type of Change

• Test update

LoopX Area

• Control plane (goals, todos, quota, scheduler, registry, runtime)

Technical Direction

• Direction / acceptance reference, when applicable: N/A; this is a test-only correction to existing Todo revision behavior.

Shared-authority RFC fixture impact

• Production-scale fixture schema: N/A
• Semantic dimensions changed, or reviewed no-impact rationale: N/A; no authority schema changed.
• Provider conformance arms run: N/A
• Read-only legacy/file/PostgreSQL three-arm rehearsal: N/A

Boundary Checklist

• No private state, credentials, raw traces, internal links, or local machine paths are included.
• No maintainer-owned benchmark work is duplicated.
• The change is scoped to this test correction.
• UI impact is marked none.
• Every commit includes a DCO Signed-off-by trailer.

[loopx-agent]

Reviewer: model_agent; model=gpt-6.1-sol; provider=OpenAI; runtime_reported; reasoning_effort=xhigh.

Reviewed exact head 3b4896427c13158e2d775de14669aae406ac4ea2 against immutable base 3ed5d6bc88ffaaa05f2b1bc6a98ed5d921379fd1. APPROVE，无阻塞发现。

动机

维护者修改 Todo 兼容入口后，需要用既有真实 provider 回归检查完整记录是否保留。
旧测试把编辑前的正文摘要带入编辑后的预期记录，正确编辑也会失败；当前测试重新计算新正文摘要，并要求它与旧摘要不同。
同一不可变基线复现旧摘要断言失败；当前精确 head 的完整测试文件 38 项通过，复杂记录与相邻任务的原有完整相等断言仍保留。
本 PR 只修正既有测试，不改变生产 Todo、权限、存储格式或 CLI/UI/Lark 行为，也不验收所有 provider 或平台。

完整读取已合并的 #5911 及基线全文摘要 helper 后判断：这是维护已有正文新鲜度契约的测试预期，不是新增 canonical 持久化字段。没有定义此断言修复的独立书面规格，因此不虚构 RFC 验收编号。

改动思路

继续使用同一个真实 canonical File fixture、实际 CLI 和完整记录相等断言。预期摘要取自声明的新正文，随后明确要求新旧摘要不同；直接从实际读回取预期值或删掉摘要比较都会削弱回归。复用现有规范化摘要 owner，未增加第二个生产决策源或单独 smoke。

具体改动

整个 PR 只有 tests/control_plane/test_local_coordination_authority.py，+9/-1：导入 todo_text_content_revision，定义 edited_text 和预期新摘要，检查非空，在完整 edited-record 预期中覆盖摘要，再断言它不同于 claimed_item 的旧摘要。原 claim、dry-run/apply、原 Markdown 移除后真实 canonical 读回、复杂 metadata、相邻 complex/successor 完整相等及 timestamp 检查全部保留。

独立对照：基线运行同一 focused function，1 failed/37 deselected，17 个字段相同，只有旧摘要与新摘要不匹配。当前 head 完整测试文件 38 passed。另用独立 hashlib 对声明新正文执行规范化并计算 SHA256，得到 3b61f6402feb68632608a56817e5dd4ffdf0b041d60b0d1d40a1d9f0a4305036，与 helper 一致且不同于旧值；预期没有从 provider 实际输出倒推。

对主干的风险

主要风险是修测试时悄悄放宽完整记录保留条件。全量 diff 和实际 fixture 执行表明，此改动只修正本应变化的派生正文摘要，仍比较其他字段及相邻任务，新旧不等检查也会拒绝陈旧摘要。Ruff、diff check、DCO 通过；未查询或等待 CI。

现有覆盖扫描区分上游 selection/concurrency 摘要测试与这个下游完整记录兼容 fixture；同作者的 hook/successor/report/process 测试 PR 维护不同边界，本 PR 没有新增同形 smoke。验证使用合成数据上的真实 File provider/CLI，不冒充 SQLite、PostgreSQL、所有平台或安装 App 验收；生产 TypeScript/权限/存储规则和用户入口没有改动。

我的整体评价

APPROVE。基线反例证明旧预期错误，当前完整文件和独立摘要 oracle 支持这个有界维护结果，没有弱化原有数据保留断言。未来重构检查已考虑 fixture 与摘要 owner：复用当前 helper 和既有测试足够，无需追加框架。评审结论不授予 merge authority。

English verdict: APPROVE — exact head 3b48964; fixes the stale expected text revision without weakening full canonical-record or sibling preservation. Immutable-base failure, 38 current-head tests, an independent normalized-text SHA256 oracle and focused static/DCO checks support this test-only maintenance change.

[mikamikasuki]

CI attribution for run 37847658548 at head 3b48964: JUnit artifact python-junit-2 reports three failing tests. I reran those exact selectors on this PR's base 3ed5d6b; all three fail there as well (3 failed in 1.60s). They are in test files already changed by open PRs #5986 (test_todo_complete_settlement_capabilities.py), #5993 (test_doubao_model_behavior_actor.py), and #5980 (test_chat_machine_configuration_api.py). This PR changes only test_local_coordination_authority.py, and its focused regression passed locally. The remaining test shards are still running, so this is an interim attribution rather than a terminal CI result.