test(todos): align thin exact-read contract

[mikamikasuki]

Goal And Delivered Outcome

• Outcome basis: Current main has a failing exact-Todo integration test and a protocol statement that contradicts the implemented exact-read contract.
• Goal/source and gap: Exact reads must preserve the complete accepted Todo requirement. docs/reference/required-work-context.md, loopx/control_plane/todos/list_readback.py, and tests/control_plane/test_todo_exact_detail.py specify that --thin is for bounded inventory lists and is rejected with --todo-id; the older integration test and active-state projection protocol had not been updated.
• Observable before → after: On main 159f00fc8ababa088e8f6cde3ef06653bc4c0511, the File and SQLite cases both fail because they expect todo list --todo-id ID --thin to return a truncated record. The integration test now checks the documented rejection, and the protocol text no longer claims exact IDs support --thin. Runtime behavior is unchanged.
• Issue/task and intended base: No separate issue is required for this self-contained maintenance correction under CONTRIBUTING.md. Intended base: main at 159f00fc8ababa088e8f6cde3ef06653bc4c0511.

Author Declaration

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

Implemented against

• Specification and revision: docs/reference/required-work-context.md and tests/control_plane/test_todo_exact_detail.py at 159f00fc8ababa088e8f6cde3ef06653bc4c0511; exact-read validation in loopx/control_plane/todos/list_readback.py.
• Criteria:

Criterion	Disposition	Symbol / path	Test or command
Exact Todo reads retain complete requirements and reject --thin	implemented	tests/test_chat_todo_detail.py	tests/test_chat_todo_detail.py
Protocol documentation matches the exact-read contract	implemented	docs/reference/protocols/active-state-structured-projection-v0.md	loopx check public-boundary scan

• Self-check before submission: Compared the failing test on the exact base and PR fix(quota): fence selected todo content freshness #5911 head, reproduced the same two provider failures on current main, read the owning CLI validation and existing exact-detail regression, searched current issues and PRs for this contract mismatch, and reviewed the two-file diff. No runtime behavior was changed.

Scope And Continuation

• Completed scope and remaining work: The integration test and protocol documentation now agree with the current exact-read contract. This resolves the two stale File/SQLite test cases without changing CLI behavior.
• Slice boundary / successor: Complete within this scope.

Validation

• Tested revision: d8fc48d50758bc54d3a08b0cb6f7c99a19767d06 (head), based on 159f00fc8ababa088e8f6cde3ef06653bc4c0511.
• Run state: finished
• Input classes: synthetic

Check kind	Result	Public-safe evidence / limitation
regression_parity	passed	The original focused test failed for both File and SQLite on 0cd547b, on current main 159f00fc, and on PR #5911 head. After the correction, all three focused test modules pass.
unit	passed	On final head d8fc48d, pytest -q tests/test_chat_todo_detail.py tests/control_plane/test_todo_exact_detail.py tests/control_plane/test_todo_projection_recovery.py: 27 passed.
static	passed	Ruff for tests/test_chat_todo_detail.py, Python compile, and git diff --check passed.
integration	passed	On code/test head 157d39a, pre-merge canary selected 12 checks, all passed, including Todo contract/read-model/projection smokes and public-boundary scan.
final-head	passed	After the protocol wording correction on d8fc48d, the three focused test modules passed (27 tests) and git diff --check passed.

• Coverage and gaps: This verifies CLI exact-read contract over both File and SQLite fixtures and aligns the protocol statement. The change does not alter runtime implementation or claim that unrelated failures in the broader PR fix(quota): fence selected todo content freshness #5911 suite are resolved.

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

• Bug fix
• New feature
• Breaking change
• Refactoring (no functional changes)
• Documentation update
• Test update

LoopX Area

• Control plane (goals, todos, quota, scheduler, registry, runtime)
• Benchmark boundary (adapters, runners, verifiers, scoring, evidence)
• Capability or extension (providers, adapters, skills)
• Public docs or presentation surface (README, protocols, dashboard)
• Build, packaging, installer, or CI
• Host or runtime integration

Technical Direction

• Direction / acceptance reference, when applicable: N/A; this is a test and protocol-documentation correction to the existing exact-read contract.

Shared-authority RFC fixture impact

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

Boundary Checklist

• Neither the diff nor this PR body/comments/attachments disclose private state, credentials, raw traces or verifier output, internal links, or local machine paths.
• I did not duplicate maintainer-owned benchmark work.
• I kept the change scoped to this self-contained test and protocol correction.
• I completed the visual evidence section for UI changes, or marked UI impact none.
• Every commit includes a DCO Signed-off-by trailer.

[mikamikasuki]

CI comparison on exact PR head d8fc48d50758bc54d3a08b0cb6f7c99a19767d06 versus the completed exact-base main run at 159f00fc8ababa088e8f6cde3ef06653bc4c0511: I compared failed test IDs from the Python shard logs (37708133072 vs 37706654513). All 29 unique shard failure IDs overlap; there are no PR-only failure IDs. The base run has one additional failure, tests/test_chat_todo_detail.py::test_exact_task_cli_http_preserve_full_current_and_retained_request[sqlite], which passes on this PR head. The remaining shard failures are present on exact-base main as well; the aggregate CI remains red due to those existing failures.

[loopx-agent]

Reviewer: model_agent | gpt-6.1-sol | OpenAI | runtime_reported | reasoning_effort=xhigh

Exact head: 029ca8c; base: 44931b6.
未发现阻塞问题 / No blocking finding.

动机

读取完整任务要求的 CLI 用户，以及维护 CLI 与聊天详情集成测试的开发者。 用户按旧协议给精确任务读取加上摘要选项时会被拒绝，开发者运行集成测试也会因相反的旧预期失败；本次把测试和双语协议改为现有的完整读取与明确拒绝规则。 独立验证表明，正常精确读取仍保留完整任务及末尾要求，摘要与精确身份同用时返回拒绝；现有集成测试继续验证聊天详情、来源恢复和读取不写状态。 本次只修正测试与协议表述，不修改运行时、存储格式、权限、用户入口或默认行为，也不承诺解决其他 PR 或整套远程 CI 的失败。

改动思路

修正已有断言和误导性协议，复用原有集成测试；不把完整需求保护改成截断。只删断言会让测试变绿却失去保护，改变运行时又会违背已接受契约。这份完整 diff 的边界合适。

规格依据：docs/reference/required-work-context.md，不可变版本 44931b6d22a50b949d43354e6ea498fb6b68d231。它在本 PR 之前已明确完整精确读取、库存概览与 --thin 拒绝规则；被修改的协议不能作为自身正确性的依据。该文档没有编号，以下是本次映射标识：

• EXACT-FULL：精确读取完整正文；当前和归档记录、末尾要求均保留。
• INVENTORY-THIN：概览可截短，精确身份与 --thin 同用必须拒绝。
• SOURCE-AUTHORITY：使用真实权威源，缺失展示和来源故障不能回退旧摘要。
• READ-ONLY：读取不提供执行授权、不写 provider 或重建缺失展示。

具体改动

tests/test_chat_todo_detail.py:43-47 的局部 CLI helper 增加 expected_exit=0；正向命令仍必须成功，负向调用明确要求 exit1。:57-58 不再期望最多500字的成功结果，而是验证拒绝及 remove --thin remedy。完整正文、末尾、真实 HTTP、归档身份、跨来源/异常来源恢复以及最终 provider revision/records 不变的断言均保留。

docs/reference/protocols/active-state-structured-projection-v0.md:229-239,253-258 同步修正英文和中文，区分概览与精确读，不引入开关或新权限。真实链是 CLI main → list_goal_todos → 既有 todo.context.page；Chat detail 使用原入口。完整生产代码与 base 相同。

重复覆盖检查已读取 exact-detail、projection-recovery 和 thin-list 既有测试；本 PR 没有新建同形 smoke。最新20个同作者 PR 及当前完整队列已检查，相关 #5911/#5882 改动投影其他路径，#5925/#5927/#5930 分属其他测试边界；本次不存在已确认的重复警告违规。

对主干的风险

最值得防的是“为了绿灯接受截断或去掉负例”。独立执行同一命令 uv run --extra test python -m pytest -q tests/test_chat_todo_detail.py tests/control_plane/test_todo_exact_detail.py tests/control_plane/test_todo_projection_recovery.py：不可变 base 25 passed、2 failed，两个失败均是旧 exit0 预期与实际拒绝冲突；exact head 27 passed，真实 File/SQLite、CLI subprocess、loopback HTTP 及既有 legacy/source 恢复覆盖通过。

另以私有隔离 mutation 让错误的精确 --thin 返回500字成功摘要，两个修改后 provider 用例均因 0 != 1 失败。它证明负例能捕获错误，不被计作产品失败；正常运行和原断言完整保留。没有用 mock 提供待证明的 HTTP/provider 成功。

原生 premerge 按完整两文件 diff 选出的12项（3个 Todo/投影、8个 docs/content、1个公开边界）及4个直接检查全部通过；Ruff、diff check、compile 通过。语义 advisory 无支持的新词汇载体，完整 diff 也未新增状态规则；空 advisory 本身不是语义证明。未查询、轮询或等待 CI；作者旧 head 的 CI 叙述未作为本次证据。未跑 packaged App、PostgreSQL、全树测试或持续负载；没有改这些边界，也不宣称它们验收通过。

typed-state、domain-neutrality、behavior-disclosure、guidance/obligation、default-off 和 authority lenses：无生产状态/命名/权限/默认/自动加载提示变更，文档如实描述现有拒绝，未把强制规则叫作建议。回滚两份文件即可，没有状态迁移。

我的整体评价

APPROVE 此 exact head。long_horizon preserved：恢复现有集成覆盖，不缩短原请求或增加重复写。user_experience improved：协议给出符合实际的使用方法，没有额外导航、确认或设置。future-facing pass 在原 helper 与测试边界复用后无需增加运行时重构；既有 reader、typed projection 和 effect authority 不动。+14/-10 不是完成证明，真实 base/head、拒绝 mutation 和来源读回才是依据。

English verdict: APPROVE — 029ca8c. This is a coherent repair of an existing stale test and bilingual protocol, based on the accepted pre-change full exact-read contract. Independent base/head reproduction, real File/SQLite CLI/HTTP checks, a decisive wrong-success mutation and native premerge pass. No runtime change, remote CI claim, unrelated PR completion or merge authority is inferred.