morluto/rea

Fork 0

mirror of https://github.com/morluto/rea.git synced 2026-11-05 06:45:22 +00:00

[PR #7062] [CLOSED] test(cli): preserve literal heading regression coverage #7615

New issue

Closed

opened 2026-11-03 13:39:08 +00:00 by bsakaguchi · 0 comments

bsakaguchi commented

2026-11-03 13:39:08 +00:00

Owner

📋 Pull Request Information

Original PR: https://github.com/morluto/rea/pull/7062
Author: @ShuaixinHuang
Created: 10/25/2026
Status: ❌ Closed

Base: main ← Head: codex/fix-cli-heading-text

📝 Commits (10+)

d36d9ac fix(cli): preserve literal keys in Markdown headings
d6f9b07 test(cli): check encoded keys in Markdown capacity errors
bbfba49 test(cli): expect literal Markdown inventory headings
89648c1 test(cli): reconcile literal heading coverage with upstream
132ec89 test(ci): align shared fixtures with current admission
1b5b635 test(ci): align cleanup and registration boundary fixtures
a2161f7 test(ci): reuse upstream native artifact fixtures
5cf98e2 test(model): preserve digest identity across display paths
08caf22 test(ci): align owned launchers and IPC uncertainty
ad7a1db test(ci): align source admission and current report fixtures

📊 Changes

10 files changed (+140 additions, -13 deletions)

View changed files

📝 src/android/JadxAnalysis.ts (+1 -1)
📝 src/android/JadxSession.ts (+3 -4)
📝 src/native/android/AndroidJniLibrary.ts (+1 -1)
📝 tests/acceptance/cliFormattingCapacity.test.ts (+4 -0)
📝 tests/acceptance/setup/hermesClientSetup.test.ts (+1 -1)
📝 tests/boundary/android/androidJniLibraries.test.ts (+1 -1)
📝 tests/boundary/cli/analysisViewCli.test.ts (+1 -1)
📝 tests/boundary/cli/markdownNestedFacts.test.ts (+65 -1)
📝 tests/boundary/filesystem/javascript-applications/javascriptApplicationEmptyLiterals.test.ts (+62 -1)
📝 tests/support/analysis-efficiency/android-jni.mjs (+1 -2)

📄 Description

Superseded by #7779 at the author's request. The replacement retains only the regression still absent from current upstream; this PR's branch and validation history are preserved.

Latest-head CI completed successfully

Verified at 2026-10-27T14:44:47.3206867Z, head 7fd6b2312a8262776535ee07fffe0ce12744c708: complete CI workflow 40518816354 and its CI required gate both finished success. This is evidence for this exact current head. Title and human author-declaration checks were also successful before this evidence-only body update; their fresh metadata checks may run again. Earlier queued/failed stages below are retained as historical records, not the current result. Local validation limitations remain disclosed. CI success does not establish maintainer approval or permission to merge.

Earlier Hermes preview verification (2026-10-27)

Current head 7fd6b2312a8262776535ee07fffe0ce12744c708 corrects the two Hermes preview assertions exposed by run 40515325267, executed as merge 361c8e0c into fixed main 79d2fc05. Upstream ddb53376 resolves existing ancestors even when the final home does not yet exist. The preview fixture now expects native sticky routing for a symlinked default root and external routing for an escaping prefix in both existence states. The test still checks both missing/existing iterations, exact configuration and skill targets, unchanged config/profile bytes, no backups and no setup writes. Relative to fixed main, this file has exactly one changed expectation line.

Windows, Node 24.22.1 / npm 11.16.0: the complete acceptance file passed 22 tests / 1 existing Windows named-user skip, including both formerly failing preview cases and all five nested-home cases. The combined two-file command recorded 52 passed / 5 existing skips / 3 failed, and is not all green: the unchanged boundary file uses three POSIX path spellings on Windows. No production identity logic or timeouts were changed to suppress these failures.

Changing the newly added upstream test block produced a real content conflict, so this branch required a normal merge of the same pinned upstream 79d2fc05; it does not follow later main changes. The resulting PR diff is five test files plus five behavior-equivalent lint normalizations required by normal hooks (duplicate imports, an array spread and redundant regex escaping). The latter were inspected; no provider capability or runtime policy is added. Normal pre-commit and pre-push typecheck/lint 2/2 passed (33.384 s), and GitHub reports no merge conflict. Latest-head CI 40518816354 has been created and was queued at verification; title and readiness succeeded. Earlier runs below are historical.

Earlier CI repair (`075fbbe4`)

Head 075fbbe41e279701b0d9a7019984309cf851728c repairs the three failures from historical run 40503239155 without changing production behavior. The Hermes positive fixtures now use the existing admitted, pinned npx launcher; anchors, aliases, comments, environment and refusal controls remain covered. The JavaScript regression validates both literal/nested entry paths within each identity namespace and follows existing application-node links with artifact digest, module and source-range provenance, instead of equating two intentionally different hash namespaces. The one upstream conflict was resolved using the maintainer-reviewed launcher assertion from fixed main 79d2fc05.

Windows / Node 24.22.1 / npm 11.16.0: complete Hermes file 12/12 passed and complete JavaScript file 25/25 passed. Normal pre-commit and pre-push typecheck/lint 2/2 passed (33.097 s). Read-only merge-tree against fixed main exited 0. These are local focused results, not a full-suite claim. Latest-head CI 40515325267 was running at verification; title and readiness succeeded. Earlier-head results below are historical.

Summary

Retain regression coverage for literal Markdown heading keys through the real CLI result-output owner after upstream 6d7e6277 implemented the production fix. Frozen input, JSON round-trip, seven exact headings, CRLF/CR injection, table output and the no-fallback path remain explicit controls.

The PR also aligns affected regression fixtures with current admission and cleanup contracts so they reach their intended failure boundaries: proven-owned stale launchers, exact retained-root paths, semantic WASM hash/format identity, the actual archive digest, and Java's standard property framing. Refusal, malformed-input, cleanup-retry and isolation assertions remain intact. The formatter itself is not overridden.

The only non-test producer correction is the existing public mobile-graph example: directory kind now drives both the root occurrence ID and its record. No provider or target-execution capability is added.

Earlier shared fixture corrections (2026-10-27)

Earlier head ad7a1db808fd39bc38f3e23021d95cf39665dc56 changes five test files only, 23 insertions / 6 deletions: a current-date PR fixture within the rolling selection window; the real ps prerequisite while retaining the empty PATH component; admitted Mach-O input for MCP inventory coverage; exact unknown-launcher diagnostics and configuration provenance; and explicit complete delivery for Android inline Evidence. Named-operation, advertised-schema/Ajv, raw Evidence, cleanup, refusal and read-only controls remain enforced. No production change or new skip is introduced.

The complete four-file Windows focused command recorded 18 passed / 15 existing POSIX skips / 1 failed. The remaining public-CLI fixture requires a POSIX fake-gh shebang; on Windows it produced gh-calls ENOENT. This command is not green. A separate probe through the real collectPrCiHealth owner with a fixed observation time confirmed old-date selection yields 0 PRs and the current-date fixture yields 1 with CI missing, complete inventories and exactly two GET calls. Ghidra, MCP inventory and launcher POSIX cases still require Linux CI.

Shared Android checks on actual failed merge 423db9e0: the existing JNI MCP owner test passed 1 selected / 4 unselected (35.66 s); that merge plus the exact four-line patch passed the complete source typecheck (17.157 s). This is shared isolated evidence; the complete POSIX Android process file was not run locally on Windows, and no real JADX engine was run.

Normal pre-commit checks for all five files and normal pre-push typecheck/lint 2/2 passed (31.057 s). As of 2026-10-27 12:23:18 UTC, this exact head's main CI 40503239155 is queued / pending, with no conclusion. Earlier green runs remain historical.

Earlier upstream compatibility

The native artifact tests now use the three canonical test/fixture files from already-merged upstream #7479 (740c0f56, inspected in main 570abe95). The inert PE bytes and manifest remain independent of the production producer. The verifier copies its relative module dependency graph while asserting that installed dependencies and packed-source inputs remain absent. Only these conflicting tests and their shared fixture were synchronized; no production parser or broad main merge was introduced.

Earlier Electron and ZCode fixture correction

Earlier head 08caf22e changed two test files only (12 insertions / 4 deletions). ZCode refresh and malformed-environment fixtures now establish owned launchers before exercising their intended admission boundary. The Electron fixture retains unresolved IPC pairing exactly: 0 paired, 7 ambiguous, 0 unpaired transmissions. Ownership, malformed-input, backup, unchanged-byte, progress and Evidence controls remain intact; this follow-up changes no production code.

Windows, Node 24.22.1 / npm 11.16.0: the complete ZCode file passed 27 tests / 1 existing Windows FIFO skip (8.23 s), and the complete Electron MCP file passed 8 tests / 2 existing Windows skips (54.01 s). Normal pre-commit checks, pre-push typecheck/lint 2/2 (34.288 s), git diff --check, and independent read-only review passed.

Historical CI run for 08caf22e used merge f8811806f2ceec7e909836d1a92f902e0310ef4a into fixed main 570abe95. Its shard 5 failed 8 cases across 4 files (2,184 passed, 30 skipped); that run's Windows lane succeeded. These earlier results do not qualify the current head.

Earlier model identity correction (`5cf98e2c`)

Historical head 5cf98e2c corrected one portable model-view regression fixture: delivery display paths may rebound while semantic artifact identity remains bound to the digest. Wrong-digest and alias-plus-wrong-digest refusals remain explicit; valid rebound paths, parent Evidence links, and cold/warm cache paths are checked.

Windows, Node 24.22.1 / npm 11.16.0: the complete focused model-identity file passed 1/1 (7.88 s). Normal pre-commit checks and pre-push typecheck/lint 2/2 passed (27.653 s).

Historical CI run for 5cf98e2c failed shard 1 on the Electron and ZCode fixture expectations corrected above (6 failed, 2,114 passed, 26 skipped). That result is superseded by the current head; it is not a current-head pass.

Earlier validation (`a2161f78`)

The preceding stage recorded 118 passed, 5 existing Windows platform skips across these unchanged boundaries:

Cleanup/WASM/TOML/output-close four-file focused command: 21 passed, 5 existing Windows platform skips. POSIX-only close-failure behavior still requires hosted Linux verification.
Complete client-environment and Java-probe files: 73 passed.
Canonical native-loader and isolated verifier files after the upstream synchronization: 24 passed.
Normal pre-commit checks and that stage's pre-push typecheck/lint 2/2 passed (25.797 s), independent read-only review, and git diff --check passed. Read-only merge-tree against fixed main 570abe95 exited 0.

Earlier validation of unchanged boundaries at the 132ec89a stage includes 66 source tests and all 9 public graph-example CLI positive/negative controls. Earlier broader local commands were not all green: a CLI command recorded a Windows temporary-directory EBUSY, another command retained an unchanged Kimi POSIX-path/Windows-separator expectation, and a full static command stopped at three upstream unused exports. These limitations are not waived or presented as successful full-suite runs. No external engine or actual Windows addon loading is claimed by the inert fixtures.

Historical hosted CI run for a2161f78 passed Source tests and the Windows lane. Shard 7 exposed the model fixture mismatch corrected above (1 failed, 2,074 passed, 34 skipped); the other seven shards were cancelled.

Author readiness

Human owner: @ShuaixinHuang

I reviewed the final diff and relevant existing code, understand the change and its edge cases, and take responsibility for the complete PR, including AI-assisted work.
I checked the evidence above, confirmed it supports the current change, and recorded meaningful validation gaps.
I declare this PR ready for initial maintainer review.

The human owner explicitly confirmed review and responsibility for this PR and requested formal maintainer review in this session on 2026-10-25.

Before merge

Concrete problem and focused conventional title.
Ordinary fix within existing scope.
Regression covers changed output and unchanged JSON/source.
No generated artifacts, secrets or unrelated cleanup.
Latest-head required CI and human readiness confirmed.

_{🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.}

## 📋 Pull Request Information **Original PR:** https://github.com/morluto/rea/pull/7062 **Author:** [@ShuaixinHuang](https://github.com/ShuaixinHuang) **Created:** 10/25/2026 **Status:** ❌ Closed **Base:** `main` ← **Head:** `codex/fix-cli-heading-text` --- ### 📝 Commits (10+) - [`d36d9ac`](https://github.com/morluto/rea/commit/d36d9ac7f1504a90f47143f3cf7f3fd11136d2eb) fix(cli): preserve literal keys in Markdown headings - [`d6f9b07`](https://github.com/morluto/rea/commit/d6f9b072c8a2dfbca83c5b764145c4152c86d692) test(cli): check encoded keys in Markdown capacity errors - [`bbfba49`](https://github.com/morluto/rea/commit/bbfba491454c42a475e2d536ff53f621994544c2) test(cli): expect literal Markdown inventory headings - [`89648c1`](https://github.com/morluto/rea/commit/89648c1e2747f54fff99ebdd0c9409ced00be9d8) test(cli): reconcile literal heading coverage with upstream - [`132ec89`](https://github.com/morluto/rea/commit/132ec89a88b76a34f221282431d0ab0584881feb) test(ci): align shared fixtures with current admission - [`1b5b635`](https://github.com/morluto/rea/commit/1b5b635b75c97f70e88d72ec5a9c7c17f6dc180c) test(ci): align cleanup and registration boundary fixtures - [`a2161f7`](https://github.com/morluto/rea/commit/a2161f78af7f8cdeb7a888b3e5bcd2f9e429f924) test(ci): reuse upstream native artifact fixtures - [`5cf98e2`](https://github.com/morluto/rea/commit/5cf98e2cae2e7a9b38de071227069f2e266dd6a3) test(model): preserve digest identity across display paths - [`08caf22`](https://github.com/morluto/rea/commit/08caf22e85ecfe60bdcc71f9272fbbc30e8ca117) test(ci): align owned launchers and IPC uncertainty - [`ad7a1db`](https://github.com/morluto/rea/commit/ad7a1db808fd39bc38f3e23021d95cf39665dc56) test(ci): align source admission and current report fixtures ### 📊 Changes **10 files changed** (+140 additions, -13 deletions) <details> <summary>View changed files</summary> 📝 `src/android/JadxAnalysis.ts` (+1 -1) 📝 `src/android/JadxSession.ts` (+3 -4) 📝 `src/native/android/AndroidJniLibrary.ts` (+1 -1) 📝 `tests/acceptance/cliFormattingCapacity.test.ts` (+4 -0) 📝 `tests/acceptance/setup/hermesClientSetup.test.ts` (+1 -1) 📝 `tests/boundary/android/androidJniLibraries.test.ts` (+1 -1) 📝 `tests/boundary/cli/analysisViewCli.test.ts` (+1 -1) 📝 `tests/boundary/cli/markdownNestedFacts.test.ts` (+65 -1) 📝 `tests/boundary/filesystem/javascript-applications/javascriptApplicationEmptyLiterals.test.ts` (+62 -1) 📝 `tests/support/analysis-efficiency/android-jni.mjs` (+1 -2) </details> ### 📄 Description Superseded by #7779 at the author's request. The replacement retains only the regression still absent from current upstream; this PR's branch and validation history are preserved. ## Latest-head CI completed successfully Verified at **2026-10-27T14:44:47.3206867Z**, head `7fd6b2312a8262776535ee07fffe0ce12744c708`: [complete CI workflow 40518816354](https://github.com/morluto/rea/actions/runs/40518816354) and its [CI required gate](https://github.com/morluto/rea/actions/runs/40518816354/job/122249198267) both finished **success**. This is evidence for this exact current head. Title and human author-declaration checks were also successful before this evidence-only body update; their fresh metadata checks may run again. Earlier queued/failed stages below are retained as historical records, not the current result. Local validation limitations remain disclosed. CI success does not establish maintainer approval or permission to merge. ## Earlier Hermes preview verification (2026-10-27) Current head `7fd6b2312a8262776535ee07fffe0ce12744c708` corrects the two Hermes preview assertions exposed by [run 40515325267](https://github.com/morluto/rea/actions/runs/40515325267/job/122228277862), executed as merge `361c8e0c` into fixed main `79d2fc05`. Upstream `ddb53376` resolves existing ancestors even when the final home does not yet exist. The preview fixture now expects native sticky routing for a symlinked default root and external routing for an escaping prefix in both existence states. The test still checks both missing/existing iterations, exact configuration and skill targets, unchanged config/profile bytes, no backups and no setup writes. Relative to fixed main, this file has exactly one changed expectation line. Windows, Node 24.22.1 / npm 11.16.0: the complete acceptance file passed **22 tests / 1 existing Windows named-user skip**, including both formerly failing preview cases and all five nested-home cases. The combined two-file command recorded **52 passed / 5 existing skips / 3 failed**, and is **not all green**: the unchanged boundary file uses three POSIX path spellings on Windows. No production identity logic or timeouts were changed to suppress these failures. Changing the newly added upstream test block produced a real content conflict, so this branch required a normal merge of the same pinned upstream `79d2fc05`; it does not follow later main changes. The resulting PR diff is five test files plus five behavior-equivalent lint normalizations required by normal hooks (duplicate imports, an array spread and redundant regex escaping). The latter were inspected; no provider capability or runtime policy is added. Normal pre-commit and pre-push typecheck/lint **2/2 passed** (33.384 s), and GitHub reports no merge conflict. [Latest-head CI 40518816354](https://github.com/morluto/rea/actions/runs/40518816354) has been created and was queued at verification; title and readiness succeeded. Earlier runs below are historical. ## Earlier CI repair (075fbbe4) Head `075fbbe41e279701b0d9a7019984309cf851728c` repairs the three failures from historical run 40503239155 without changing production behavior. The Hermes positive fixtures now use the existing admitted, pinned npx launcher; anchors, aliases, comments, environment and refusal controls remain covered. The JavaScript regression validates both literal/nested entry paths within each identity namespace and follows existing application-node links with artifact digest, module and source-range provenance, instead of equating two intentionally different hash namespaces. The one upstream conflict was resolved using the maintainer-reviewed launcher assertion from fixed main `79d2fc05`. Windows / Node 24.22.1 / npm 11.16.0: complete Hermes file **12/12 passed** and complete JavaScript file **25/25 passed**. Normal pre-commit and pre-push typecheck/lint **2/2 passed** (33.097 s). Read-only merge-tree against fixed main exited 0. These are local focused results, not a full-suite claim. [Latest-head CI 40515325267](https://github.com/morluto/rea/actions/runs/40515325267) was running at verification; title and readiness succeeded. Earlier-head results below are historical. ## Summary Retain regression coverage for literal Markdown heading keys through the real CLI result-output owner after upstream `6d7e6277` implemented the production fix. Frozen input, JSON round-trip, seven exact headings, CRLF/CR injection, table output and the no-fallback path remain explicit controls. The PR also aligns affected regression fixtures with current admission and cleanup contracts so they reach their intended failure boundaries: proven-owned stale launchers, exact retained-root paths, semantic WASM hash/format identity, the actual archive digest, and Java's standard property framing. Refusal, malformed-input, cleanup-retry and isolation assertions remain intact. The formatter itself is not overridden. The only non-test producer correction is the existing public mobile-graph example: directory kind now drives both the root occurrence ID and its record. No provider or target-execution capability is added. ## Earlier shared fixture corrections (2026-10-27) Earlier head `ad7a1db808fd39bc38f3e23021d95cf39665dc56` changes **five test files only, 23 insertions / 6 deletions**: a current-date PR fixture within the rolling selection window; the real `ps` prerequisite while retaining the empty PATH component; admitted Mach-O input for MCP inventory coverage; exact unknown-launcher diagnostics and configuration provenance; and explicit complete delivery for Android inline Evidence. Named-operation, advertised-schema/Ajv, raw Evidence, cleanup, refusal and read-only controls remain enforced. No production change or new skip is introduced. The complete four-file Windows focused command recorded **18 passed / 15 existing POSIX skips / 1 failed**. The remaining public-CLI fixture requires a POSIX fake-`gh` shebang; on Windows it produced `gh-calls` ENOENT. This command is **not green**. A separate probe through the real `collectPrCiHealth` owner with a fixed observation time confirmed old-date selection yields 0 PRs and the current-date fixture yields 1 with CI `missing`, complete inventories and exactly two GET calls. Ghidra, MCP inventory and launcher POSIX cases still require Linux CI. Shared Android checks on actual failed merge `423db9e0`: the existing JNI MCP owner test passed **1 selected / 4 unselected** (35.66 s); that merge plus the exact four-line patch passed the complete source typecheck (17.157 s). This is shared isolated evidence; the complete POSIX Android process file was not run locally on Windows, and no real JADX engine was run. Normal pre-commit checks for all five files and normal pre-push typecheck/lint **2/2 passed** (31.057 s). As of **2026-10-27 12:23:18 UTC**, this exact head's [main CI 40503239155](https://github.com/morluto/rea/actions/runs/40503239155) is **queued / pending**, with no conclusion. Earlier green runs remain historical. ## Earlier upstream compatibility The native artifact tests now use the three canonical test/fixture files from already-merged upstream #7479 (`740c0f56`, inspected in main `570abe95`). The inert PE bytes and manifest remain independent of the production producer. The verifier copies its relative module dependency graph while asserting that installed dependencies and packed-source inputs remain absent. Only these conflicting tests and their shared fixture were synchronized; no production parser or broad main merge was introduced. ## Earlier Electron and ZCode fixture correction Earlier head `08caf22e` changed two test files only (**12 insertions / 4 deletions**). ZCode refresh and malformed-environment fixtures now establish owned launchers before exercising their intended admission boundary. The Electron fixture retains unresolved IPC pairing exactly: **0 paired, 7 ambiguous, 0 unpaired** transmissions. Ownership, malformed-input, backup, unchanged-byte, progress and Evidence controls remain intact; this follow-up changes no production code. Windows, Node 24.22.1 / npm 11.16.0: the complete ZCode file passed **27 tests / 1 existing Windows FIFO skip** (8.23 s), and the complete Electron MCP file passed **8 tests / 2 existing Windows skips** (54.01 s). Normal pre-commit checks, pre-push typecheck/lint **2/2** (34.288 s), `git diff --check`, and independent read-only review passed. Historical [CI run for `08caf22e`](https://github.com/morluto/rea/actions/runs/40500791494) used merge `f8811806f2ceec7e909836d1a92f902e0310ef4a` into fixed main `570abe95`. Its [shard 5](https://github.com/morluto/rea/actions/runs/40500791494/job/122178204634) failed **8 cases across 4 files** (2,184 passed, 30 skipped); that run's Windows lane succeeded. These earlier results do not qualify the current head. ## Earlier model identity correction (`5cf98e2c`) Historical head `5cf98e2c` corrected one portable model-view regression fixture: delivery display paths may rebound while semantic artifact identity remains bound to the digest. Wrong-digest and alias-plus-wrong-digest refusals remain explicit; valid rebound paths, parent Evidence links, and cold/warm cache paths are checked. Windows, Node 24.22.1 / npm 11.16.0: the complete focused model-identity file passed **1/1** (7.88 s). Normal pre-commit checks and pre-push typecheck/lint **2/2 passed** (27.653 s). Historical [CI run for `5cf98e2c`](https://github.com/morluto/rea/actions/runs/40495696096) failed shard 1 on the Electron and ZCode fixture expectations corrected above (6 failed, 2,114 passed, 26 skipped). That result is superseded by the current head; it is not a current-head pass. ## Earlier validation (`a2161f78`) The preceding stage recorded **118 passed, 5 existing Windows platform skips** across these unchanged boundaries: - Cleanup/WASM/TOML/output-close four-file focused command: **21 passed, 5 existing Windows platform skips**. POSIX-only close-failure behavior still requires hosted Linux verification. - Complete client-environment and Java-probe files: **73 passed**. - Canonical native-loader and isolated verifier files after the upstream synchronization: **24 passed**. - Normal pre-commit checks and that stage's pre-push typecheck/lint **2/2 passed** (25.797 s), independent read-only review, and `git diff --check` passed. Read-only `merge-tree` against fixed main `570abe95` exited 0. Earlier validation of unchanged boundaries at the `132ec89a` stage includes 66 source tests and all 9 public graph-example CLI positive/negative controls. Earlier broader local commands were not all green: a CLI command recorded a Windows temporary-directory `EBUSY`, another command retained an unchanged Kimi POSIX-path/Windows-separator expectation, and a full static command stopped at three upstream unused exports. These limitations are not waived or presented as successful full-suite runs. No external engine or actual Windows addon loading is claimed by the inert fixtures. Historical hosted [CI run for `a2161f78`](https://github.com/morluto/rea/actions/runs/40493479456) passed Source tests and the Windows lane. Shard 7 exposed the model fixture mismatch corrected above (1 failed, 2,074 passed, 34 skipped); the other seven shards were cancelled. ## Author readiness Human owner: @ShuaixinHuang - [x] I reviewed the final diff and relevant existing code, understand the change and its edge cases, and take responsibility for the complete PR, including AI-assisted work. - [x] I checked the evidence above, confirmed it supports the current change, and recorded meaningful validation gaps. - [x] I declare this PR ready for initial maintainer review. The human owner explicitly confirmed review and responsibility for this PR and requested formal maintainer review in this session on 2026-10-25. ## Before merge - [x] Concrete problem and focused conventional title. - [x] Ordinary fix within existing scope. - [x] Regression covers changed output and unchanged JSON/source. - [x] No generated artifacts, secrets or unrelated cleanup. - [ ] Latest-head required CI and human readiness confirmed. --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>

bsakaguchi

2026-11-03 13:39:08 +00:00

closed this issue
added the
pull-request
label

No milestone

No project

No assignees

1 participant

Notifications

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference

morluto/rea#7615

No description provided.

Rows
Columns