test(architecture): pin result files digest consumer

[mikamikasuki]

Goal And Delivered Outcome

• Outcome basis: A reproducible current-main CI failure in the content-digest consumer ownership test.
• Goal/source and gap: The test pins every Python module that delegates stored SHA-256 matching to the canonical content-digest owner. Merged commit 20d16ee added loopx.control_plane.collaboration.result_files, which imports the canonical BARE_SHA256_PATTERN, but the pinned consumer tuple did not include it.
• Observable before → after: At main 159f00f, the exact test fails because the derived consumer set contains loopx.control_plane.collaboration.result_files and the pinned set does not. This change adds that consumer to the pin; the full focused module then passes (211 passed).
• Issue/task and intended base: No separate issue is required for this self-contained reproduced defect under CONTRIBUTING.md#find-work. Base main at 159f00f.

Author Declaration

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

Implemented against

• Specification and revision: The consumer-ownership invariant documented in tests/architecture/test_content_digest_single_owner.py at 159f00f.
• Criteria:

Criterion	Disposition	Symbol / path	Test or command
Every module importing the canonical digest matcher is present in the reviewed pinned set	implemented	CONSUMER_MODULES	tests/architecture/test_content_digest_single_owner.py

• Self-check before submission: Read the contribution, governance, issue and PR guidance; inspected the introducing commit and current consumer; searched issues and PRs for a duplicate; reproduced the failure on exact main; confirmed the change adds only the missing pinned consumer.

Scope And Continuation

• Completed scope and remaining work: The current-main architecture manifest regression is fixed. No production code or digest behavior changes.
• Slice boundary / successor: Complete within this scope; no successor is needed.

Validation

• Tested revision: a058001 (base 159f00f).
• Run state: finished.
• Input classes: none.

Check kind	Result	Public-safe evidence
regression_parity	passed	On exact base, the targeted consumer-manifest test fails with result_files missing from the pinned set. On this head, tests/architecture/test_content_digest_single_owner.py passes: 211 passed.
static	passed	Ruff check and Python compile passed; git diff --check passed.
integration	passed	loopx canary premerge --from-git-diff: 2 selected, 2 passed, zero failures and zero manual holds.

• Coverage and gaps: The change only updates a static ownership manifest; the existing architecture test verifies the manifest against all repository consumers. No runtime behavior changes, so runtime tests are not applicable. GitHub CI remains pending.

Frontend / Visual Evidence

• UI impact: none
• Before: N/A
• After: N/A
• States and viewports shown: N/A
• Source data: none
• 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 focused architecture-test manifest repair.

Shared-authority RFC fixture impact

• Production-scale fixture schema: N/A
• Semantic dimensions changed, or reviewed no-impact rationale: N/A; the test-only pin does not alter authority schemas or runtime semantics.
• Provider conformance arms run: N/A
• Read-only legacy/file/PostgreSQL three-arm rehearsal: N/A

Boundary Checklist

• No private state, credentials, raw traces, verifier output, internal links or local machine paths are included.
• This does not duplicate maintainer-owned benchmark work.
• The change is scoped to the reproduced consumer-manifest failure.
• UI impact is none.
• The commit includes a DCO Signed-off-by trailer.

[loopx-agent]

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

Exact head: a058001

动机

运行摘要所有权测试的贡献者，需要区分真实的校验规则漂移与遗漏登记造成的误报。 主干新增附件文件读写模块后，该模块已经复用统一 SHA-256 校验器，但测试的已审消费者清单缺了一项；补入真实消费者后，误报消失，复制校验器仍会被拒绝。 独立对照确认基线唯一失败是 result_files 未登记；当前 head 的 253 项相关测试通过，且两种复制校验器的反例都被现有所有权测试捕获。 本 PR 只维护既有测试清单，不改变附件协议、运行时摘要规则、权限、默认配置或实际 Lark 回传行为。

改动思路

沿用已有“显式审阅清单 + 独立源码扫描”的架构守卫。保持清单不动会留下已复现的主干误报；删除集合相等断言，或从扫描结果自动生成期望清单，会取消新消费者必须被审阅的边界。补入已经真实使用统一校验器的模块，是最小且完整的维护修复。

规格依据：tests/architecture/test_content_digest_single_owner.py，不可变版本 159f00fc8ababa088e8f6cde3ef06653bc4c0511；先读原有模块说明及断言，再检查本次 diff。原规格的四层互不替代，本次只补齐消费者登记，保留形状扫描、canonical export 引用、对象身份和可读性检查。

• test_the_consumer_manifest_is_exactly_the_pinned_set：当前 head 已实现并通过。
• test_consumers_name_the_canonical_exports_and_use_them：当前 head 已实现并通过。
• test_every_consumer_holds_the_owner_object_not_an_equal_copy：当前 head 已实现并通过。
• test_no_consumer_holds_a_regex_this_file_cannot_read：当前 head 已实现并通过。

具体改动

唯一改动是上述测试文件第184行：向 CONSUMER_MODULES 增加 loopx.control_plane.collaboration.result_files，一行新增、零删除。实际模块第12行已从既有 content_digest 导入 BARE_SHA256_PATTERN，第21行的 result_file_path 使用该对象的 fullmatch；本 PR 没有新增摘要规则或生产调用方。

正向：普通 pytest 入口 → _repo_scan 推导真实 import → 显式清单相等 → 原有 export/use、对象身份和可读性检查均纳入这个消费者。独立重跑两份架构模块，基线252通过/1失败，head253通过；唯一原失败准确指向未登记的 result_files。

反向：在独立进程中的真实模块上注入同形状、不同身份的 regex，再运行原有对象身份测试；literal 与运行时拼接两种复制在基线均漏检，在当前 head 均被拒绝。恢复后通过，额外虚构 pin 仍被拒绝。同一冻结脚本还进入真实摘要路径及文件读回，验证大写、短/长串、非十六进制、带 envelope、换行、空/null 值，尺寸不符和同尺寸错误字节仍拒绝。共18组 head全部满足预期，基线15/18；3项差异正是本次修复的清单与覆盖缺口。两版实际生产模块字节一致，没有 mock 提供所测结论，没有改写仓库或活动数据。

对主干的风险

没有阻塞发现。风险是把补清单误当成放宽断言；完整 diff 证实只新增真实消费者，原断言没有弱化。源码及既有覆盖搜索确认仍复用同一测试 owner，另一生产所有权测试覆盖的是构造边界，无法替代这个显式清单。已检查同作者30项近期 PR 元数据，未发现本项重复清单修复或同形状临时示例批次；不同边界的维护修复不因同作者而被否定。

标准 premerge 的4项直接检查、2项 catalog canary（maintainability/full semantic vocabulary）通过，无失败、跳过或 manual hold；advisory 没有支持的新增 vocabulary carrier，但测试路径/动态语义不在其保证内。遵守 wait_for_ci=false，未查询、轮询或等待 CI。首次覆盖搜索使用不存在的测试路径，已改为真实 architecture/extensions/control_plane 路径并成功重跑；这是检查命令错误，不是 PR 缺陷。

未验证 live Lark、跨宿主服务恢复或完整用户业务流程；一行测试登记不改变这些行为，也不宣称它们完成。静态扫描原有动态构造/数据输入等覆盖限制保留。回滚是撤回单一登记；未来消费者移除或增加仍需显式审阅，不能自动推导期望集合来隐藏 fan-out 变化。

我的整体评价

APPROVE 当前精确 head。它关闭的是可复现的既有测试清单遗漏，并证明新增消费者会受原有身份守卫约束，具有实际维护价值。没有新 smoke、fixture 或并行决策 owner。未来改进复核认为当前 tuple 是合适的既有归属；拆分大测试文件与本项缺口无关，不在本 PR 扩大重构。运行时、持久化状态、权限和默认入口均未改变；此结论不授予合并或发布权限。

English verdict: APPROVE - a058001. Adds the genuine existing result_files consumer to the reviewed digest-ownership pin without weakening assertions. Independent immutable-base/head validation:253 head tests passed versus one reproduced base omission;18 paired head cases include two copied-pattern mutations that the base misses. Standard premerge passed; CI not consulted. No production behavior or delivery-completion claim.

[mikamikasuki]

CI attribution update: Run 37712780676's Python JUnit artifact includes a failure for tests/control_plane/test_remote_location_shape_owner.py in the decision_context file-URL case. I reproduced it on exact current main 44931b6 with uv run --extra test python -m pytest -q tests/control_plane/test_remote_location_shape_owner.py -k decision_context: the file URL is rejected with the existing local-path message, while the s3 URL case passes. This is an existing test-expectation mismatch outside this PR's one-line consumer-manifest change. PR #5930 updates the test to assert each entry point's file-URL message separately from its remote-URL message; it changes tests only and leaves runtime behavior untouched.