feat(ci): tighten PR template + add soft-fail quality gates (#965)#1001
Merged
senamakel merged 10 commits intotinyhumansai:mainfrom Apr 29, 2026
Merged
feat(ci): tighten PR template + add soft-fail quality gates (#965)#1001senamakel merged 10 commits intotinyhumansai:mainfrom
senamakel merged 10 commits intotinyhumansai:mainfrom
Conversation
Contributor
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: defaults Review profile: CHILL Plan: Pro Run ID: 📒 Files selected for processing (1)
🚧 Files skipped from review as they are similar to previous changes (1)
📝 WalkthroughWalkthroughAdds PR template tightening, a soft-fail "PR Quality" GitHub Actions workflow, parser libraries and CLI scripts to validate PR checklists and the test coverage matrix, unit tests for parsers, and an auto-generated feature ID catalog JSON. Changes
Sequence Diagram(s)sequenceDiagram
participant Author
participant GitHub as "GitHub PR"
participant Actions as "PR Quality Workflow"
participant Checklist as "check-pr-checklist.mjs"
participant Matrix as "check-coverage-matrix.mjs"
participant Lychee as "markdown link checker"
Author->>GitHub: Open/Update PR with body
GitHub->>Actions: Trigger workflow (labels/edit)
Actions->>Checklist: Run checklist job (PR_BODY)
Checklist-->>Actions: Summary + exit code
Actions->>Matrix: Run coverage-matrix job
Matrix->>Matrix: load TEST-COVERAGE-MATRIX.md + feature-ids.json
Matrix-->>Actions: Validation report + exit code
Actions->>Lychee: Run markdown link checks
Lychee-->>Actions: Link check results
Actions-->>GitHub: Job statuses (soft-fail / continue-on-error)
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~20 minutes Possibly related PRs
Suggested reviewers
Poem
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Review rate limit: 2/5 reviews remaining, refill in 30 minutes and 57 seconds. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Actionable comments posted: 3
🧹 Nitpick comments (3)
scripts/lib/coverage-matrix-parser.mjs (1)
15-22: Inconsistent error handling between ID and status validation.Invalid IDs cause the row to be skipped (line 17:
continue), but invalid statuses still push the row torows(line 22). This means duplicate detection and catalog validation will include rows with invalid statuses. Consider whether this asymmetry is intentional.♻️ Option: Skip rows with invalid status for consistency
if (!VALID_STATUS.has(status)) { errors.push(`Row ${id}: invalid status "${status}" (must be one of ${[...VALID_STATUS].join(' ')})`); + continue; } rows.push({ id, name, layer, path, status, notes });🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@scripts/lib/coverage-matrix-parser.mjs` around lines 15 - 22, The code currently skips rows with invalid IDs (using ID_REGEX test + continue) but still pushes rows with invalid statuses into rows, causing downstream validation on bad data; update the status-validation branch around VALID_STATUS and errors so that when status is invalid you do not push the row (either add a continue after pushing the status error or move rows.push into the valid-status branch) — refer to ID_REGEX, VALID_STATUS, errors, and rows to locate and fix the logic..github/workflows/pr-quality.yml (1)
13-17: Soft-fail configuration is appropriate for the soak period.The
continue-on-error: trueon each job allows failures to surface without blocking PRs. Consider adding a comment or issue reference noting when to flip these to required status checks.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In @.github/workflows/pr-quality.yml around lines 13 - 17, The job "checklist-guard" currently sets continue-on-error: true; add a brief inline comment above or beside continue-on-error referencing the soak period end or an issue/PR number that tracks switching this to required status checks, e.g., mention the target date or issue ID to flip continue-on-error to false, so maintainers can find and update the continue-on-error setting for checklist-guard when the soak completes.scripts/check-coverage-matrix.mjs (1)
9-10: No error handling for missing files.If
docs/TEST-COVERAGE-MATRIX.mdorscripts/feature-ids.jsondoesn't exist, the script will throw an unhandled exception with a cryptic stack trace. Consider wrapping in try-catch for a friendlier CI error message.🛡️ Proposed error handling
-const matrixMd = await readFile(join(repoRoot, 'docs/TEST-COVERAGE-MATRIX.md'), 'utf8'); -const catalog = JSON.parse(await readFile(join(repoRoot, 'scripts/feature-ids.json'), 'utf8')); +let matrixMd, catalog; +try { + matrixMd = await readFile(join(repoRoot, 'docs/TEST-COVERAGE-MATRIX.md'), 'utf8'); +} catch (err) { + console.error(`Failed to read coverage matrix: ${err.message}`); + process.exit(1); +} +try { + catalog = JSON.parse(await readFile(join(repoRoot, 'scripts/feature-ids.json'), 'utf8')); +} catch (err) { + console.error(`Failed to read/parse feature-ids.json: ${err.message}`); + process.exit(1); +}🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@scripts/check-coverage-matrix.mjs` around lines 9 - 10, The two await readFile calls that populate matrixMd and catalog lack error handling, so missing files or invalid JSON will throw uncaught exceptions; wrap the file reads (and JSON.parse for catalog) in a try-catch around the code that calls readFile and JSON.parse, and on error print a clear, friendly message including the file name and the error (use the join(repoRoot, 'docs/TEST-COVERAGE-MATRIX.md') and join(repoRoot, 'scripts/feature-ids.json') paths for context), then call process.exit(1) so CI fails with a readable error; update the block that assigns matrixMd and catalog to use this try-catch and ensure JSON.parse errors are surfaced with the same friendly message.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In @.github/PULL_REQUEST_TEMPLATE.md:
- Line 24: The checklist entry in .github/PULL_REQUEST_TEMPLATE.md referencing
"docs/RELEASE-MANUAL-SMOKE.md" is broken; either add the missing file
docs/RELEASE-MANUAL-SMOKE.md with the required manual smoke checklist content or
update the checklist line in .github/PULL_REQUEST_TEMPLATE.md to reference an
existing document (e.g., replace the path or remove the checkbox) so the link no
longer causes CI to fail; ensure the updated checklist text exactly replaces the
current "- [ ] Manual smoke checklist updated if this touches release-cut
surfaces ([`docs/RELEASE-MANUAL-SMOKE.md`](../docs/RELEASE-MANUAL-SMOKE.md))"
entry.
In @.github/workflows/pr-quality.yml:
- Around line 57-66: The Lychee link check step ("Lychee link check") includes
'.github/PULL_REQUEST_TEMPLATE.md' in the args list which references a missing
file 'docs/RELEASE-MANUAL-SMOKE.md', causing the check to fail; fix by either
removing the '.github/PULL_REQUEST_TEMPLATE.md' entry from the args array in the
lychee-action step, updating the PULL_REQUEST_TEMPLATE.md to remove or correct
the broken 'docs/RELEASE-MANUAL-SMOKE.md' link, or add the missing
'docs/RELEASE-MANUAL-SMOKE.md' file so the reference resolves.
In `@scripts/lib/checklist-parser.mjs`:
- Around line 9-17: The parser currently treats every line the same and includes
checklist markers inside fenced code blocks; modify the for loop that iterates
body.split(/\r?\n/) to track an inFence boolean toggled when encountering fence
delimiters (e.g., lines that match /^\s*```/ or /^\s*~~~/) and skip processing
(continue) while inFence is true so ITEM_REGEX and NA_REGEX matching (the
existing match/checked/text/naReason logic and items.push) only runs for lines
outside fenced code blocks.
---
Nitpick comments:
In @.github/workflows/pr-quality.yml:
- Around line 13-17: The job "checklist-guard" currently sets continue-on-error:
true; add a brief inline comment above or beside continue-on-error referencing
the soak period end or an issue/PR number that tracks switching this to required
status checks, e.g., mention the target date or issue ID to flip
continue-on-error to false, so maintainers can find and update the
continue-on-error setting for checklist-guard when the soak completes.
In `@scripts/check-coverage-matrix.mjs`:
- Around line 9-10: The two await readFile calls that populate matrixMd and
catalog lack error handling, so missing files or invalid JSON will throw
uncaught exceptions; wrap the file reads (and JSON.parse for catalog) in a
try-catch around the code that calls readFile and JSON.parse, and on error print
a clear, friendly message including the file name and the error (use the
join(repoRoot, 'docs/TEST-COVERAGE-MATRIX.md') and join(repoRoot,
'scripts/feature-ids.json') paths for context), then call process.exit(1) so CI
fails with a readable error; update the block that assigns matrixMd and catalog
to use this try-catch and ensure JSON.parse errors are surfaced with the same
friendly message.
In `@scripts/lib/coverage-matrix-parser.mjs`:
- Around line 15-22: The code currently skips rows with invalid IDs (using
ID_REGEX test + continue) but still pushes rows with invalid statuses into rows,
causing downstream validation on bad data; update the status-validation branch
around VALID_STATUS and errors so that when status is invalid you do not push
the row (either add a continue after pushing the status error or move rows.push
into the valid-status branch) — refer to ID_REGEX, VALID_STATUS, errors, and
rows to locate and fix the logic.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: 033986f8-a05b-492b-a41c-437772c26756
📒 Files selected for processing (9)
.github/PULL_REQUEST_TEMPLATE.md.github/workflows/pr-quality.ymlapp/test/scripts/checklist-parser.test.tsapp/test/scripts/coverage-matrix-parser.test.tsscripts/check-coverage-matrix.mjsscripts/check-pr-checklist.mjsscripts/feature-ids.jsonscripts/lib/checklist-parser.mjsscripts/lib/coverage-matrix-parser.mjs
Comment on lines
+57
to
+66
| - name: Lychee link check | ||
| uses: lycheeverse/lychee-action@v2 | ||
| with: | ||
| args: >- | ||
| --no-progress | ||
| --include-fragments | ||
| 'docs/**/*.md' | ||
| 'src/**/README.md' | ||
| '.github/PULL_REQUEST_TEMPLATE.md' | ||
| fail: true |
Contributor
There was a problem hiding this comment.
🧩 Analysis chain
🏁 Script executed:
#!/bin/bash
# Verify the referenced file doesn't exist
ls -la docs/RELEASE-MANUAL-SMOKE.md 2>&1 || echo "File not found (expected)"Repository: tinyhumansai/openhuman
Length of output: 168
🏁 Script executed:
find . -name "PULL_REQUEST_TEMPLATE.md" -o -name "PULL_REQUEST_TEMPLATE*" | head -20Repository: tinyhumansai/openhuman
Length of output: 101
🏁 Script executed:
cat .github/PULL_REQUEST_TEMPLATE.md 2>&1 || echo "File not found"Repository: tinyhumansai/openhuman
Length of output: 1846
Link check includes the template with a broken link.
The glob on line 65 includes .github/PULL_REQUEST_TEMPLATE.md, which contains a reference to docs/RELEASE-MANUAL-SMOKE.md — a file that doesn't exist. The lychee check will fail until that file is created or the reference is removed.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In @.github/workflows/pr-quality.yml around lines 57 - 66, The Lychee link check
step ("Lychee link check") includes '.github/PULL_REQUEST_TEMPLATE.md' in the
args list which references a missing file 'docs/RELEASE-MANUAL-SMOKE.md',
causing the check to fail; fix by either removing the
'.github/PULL_REQUEST_TEMPLATE.md' entry from the args array in the
lychee-action step, updating the PULL_REQUEST_TEMPLATE.md to remove or correct
the broken 'docs/RELEASE-MANUAL-SMOKE.md' link, or add the missing
'docs/RELEASE-MANUAL-SMOKE.md' file so the reference resolves.
oxoxDev
added a commit
to oxoxDev/openhuman
that referenced
this pull request
Apr 28, 2026
…#965) - checklist-parser: skip checklist markers inside fenced code blocks (CodeRabbit found `- [ ]` lines inside ``` fences would falsely match). - checklist-parser: tighten N/A regex to `(?:\(N\/A\)|N\/A\b)` so `(N/A) reason` correctly consumes the closing paren and `N/Always` doesn't false-positive. - coverage-matrix-parser: drop rows with invalid status from `rows[]` (was asymmetric — invalid IDs `continue`d but invalid statuses still pushed). Fixes double-counting downstream. - check-coverage-matrix: wrap input reads in try/catch so missing files surface a clear message instead of an unhandled rejection. - pr-quality.yml: comment the 2-week soft-fail soak intent on the `continue-on-error: true` block. - Move vitest specs from app/test/scripts/ to app/test/ — vitest config only includes `test/*.test.{ts,tsx}` (not nested), so the scripts/ subdir was silently never running. Two cases added in this round (fenced-block skip + invalid-status drop) bring the total to 9. Refs tinyhumansai#965 Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Replaces the existing 5-item checklist in .github/PULL_REQUEST_TEMPLATE.md with a 6-item schema-locked version covering test additions (with failure-path requirement), coverage matrix sync, feature-ID listing, mock-policy compliance, manual-smoke updates, and Closes-keyword usage. N/A guidance is mandatory — items must be marked N/A with a one-line reason rather than deleted. Refs tinyhumansai#965 Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds two pure-Node-stdlib parsers under scripts/lib/:
- checklist-parser.mjs: parses GitHub markdown PR-body checklists into
{checked, naReason, text} items, treating "N/A: <reason>" as
satisfied even when unchecked.
- coverage-matrix-parser.mjs: parses docs/TEST-COVERAGE-MATRIX.md rows,
validates IDs against `^\d+(?:\.\d+){2,3}$` (covers 3- and
4-component IDs — section 3.3 uses 4-component), validates status
legend (✅/🟡/❌/🚫), and reports duplicates + catalog gaps.
Refs tinyhumansai#965
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- scripts/check-pr-checklist.mjs reads PR_BODY env, parses, exits non-zero if any non-N/A item is unchecked. - scripts/check-coverage-matrix.mjs reads docs/TEST-COVERAGE-MATRIX.md + scripts/feature-ids.json and validates parse errors, missing IDs, and duplicates. - scripts/feature-ids.json catalogs all 136 feature IDs sorted numerically (3- and 4-component, derived from the matrix). Acts as the source-of-truth for the matrix-sync gate. Refs tinyhumansai#965 Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- app/test/scripts/checklist-parser.test.ts (5 cases): empty body, all-checked, unchecked-no-NA, mix with N/A reasons, mixed-case [x]/[X]. - app/test/scripts/coverage-matrix-parser.test.ts (2 cases): three valid rows including a 4-component ID; duplicate-ID + invalid-status + missing-from-catalog detection. Imports use `@ts-expect-error` for the relative .mjs import since the parser modules under scripts/lib/ are pure ESM with no .d.ts. Refs tinyhumansai#965 Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
…i#965) Adds .github/workflows/pr-quality.yml with three jobs all running on ubuntu-latest, all `continue-on-error: true` for a 2-week soft-fail window, all skipped when the PR has the `docs` or `chore` label: - checklist-guard: runs scripts/check-pr-checklist.mjs against the PR body via PR_BODY env. - coverage-matrix: runs scripts/check-coverage-matrix.mjs against the matrix + feature-ids catalog. - markdown-link-check: lycheeverse/lychee-action@v2 with glob `'docs/**/*.md' 'src/**/README.md' '.github/PULL_REQUEST_TEMPLATE.md'` so tinyhumansai#966 domain READMEs inherit the gate. Concurrency cancels superseded runs by PR number / head ref. Closes tinyhumansai#965 Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Removes single-arg arrow-fn parens and adds blank-line after `vitest` import per project prettier config. No behaviour change. Refs tinyhumansai#965 Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
…#965) - checklist-parser: skip checklist markers inside fenced code blocks (CodeRabbit found `- [ ]` lines inside ``` fences would falsely match). - checklist-parser: tighten N/A regex to `(?:\(N\/A\)|N\/A\b)` so `(N/A) reason` correctly consumes the closing paren and `N/Always` doesn't false-positive. - coverage-matrix-parser: drop rows with invalid status from `rows[]` (was asymmetric — invalid IDs `continue`d but invalid statuses still pushed). Fixes double-counting downstream. - check-coverage-matrix: wrap input reads in try/catch so missing files surface a clear message instead of an unhandled rejection. - pr-quality.yml: comment the 2-week soft-fail soak intent on the `continue-on-error: true` block. - Move vitest specs from app/test/scripts/ to app/test/ — vitest config only includes `test/*.test.{ts,tsx}` (not nested), so the scripts/ subdir was silently never running. Two cases added in this round (fenced-block skip + invalid-status drop) bring the total to 9. Refs tinyhumansai#965 Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Drop unused @ts-expect-error directives and add explicit param types for .map / .every / .some callbacks. .mjs imports resolve through ts module resolution without the directive; CI tsc --noEmit was failing on TS2578 + TS7006 even though local somehow tolerated it. Refs tinyhumansai#965
oxoxDev
force-pushed
the
feat/965-pr-template-ci-guard
branch
from
42a53ee to
6152b91
Compare
Contributor
Author
|
@senamakel addressed:
Local gates green: CI re-running on |
…ai#965) Markdown Link Check was hard-blocking PR merge on 2 pre-existing main errors unrelated to this PR: - docs/install.md#apt-debianubuntu — slug mismatch with lychee's fragment resolver (heading text contains parens + slash) - http://localhost:9222 in docs/webview-integration-playbook.md — a legit dev-only URL that should never be probed by CI Add narrow excludes so the gate flags only new dead links introduced by future PRs. Branch protection will unblock this PR; pre-existing errors can be cleaned up in a follow-up rather than scope-creeping this one. Refs tinyhumansai#965
Contributor
There was a problem hiding this comment.
Actionable comments posted: 1
🧹 Nitpick comments (1)
app/test/coverage-matrix-parser.test.ts (1)
22-40: Add a malformed-ID regression test.Please add a case asserting that a row like
| 1.2 | ... |yields anInvalid ID formaterror. This will lock the intended ID-validation contract and prevent silent parser regressions.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@app/test/coverage-matrix-parser.test.ts` around lines 22 - 40, Add a new test that feeds parseMatrix a row with a malformed ID like `| 1.2 | ... |` and asserts the parser emits an "Invalid ID format" error; specifically call parseMatrix(md) where md contains the `1.2` row and assert parsed.errors.some(e => e.includes('Invalid ID format')) is true (and optionally assert the malformed row is not present in parsed.rows). Use the existing test style and functions (parseMatrix, parsed.errors, parsed.rows) to locate where to add this new case.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@scripts/lib/coverage-matrix-parser.mjs`:
- Around line 1-16: ROW_REGEX is too strict so malformed IDs don't match the row
pattern and never hit the ID_REGEX validation in parseMatrix; change ROW_REGEX
(used by parseMatrix) to capture the first table cell generically (e.g. accept
any non-pipe content) instead of enforcing the dotted-number shape, then keep
the existing ID_REGEX check to push `Invalid ID format: ${id}` into errors when
the captured id fails validation; update the ROW_REGEX constant and ensure
parseMatrix still trims and validates the captured id against ID_REGEX and
records errors for bad IDs.
---
Nitpick comments:
In `@app/test/coverage-matrix-parser.test.ts`:
- Around line 22-40: Add a new test that feeds parseMatrix a row with a
malformed ID like `| 1.2 | ... |` and asserts the parser emits an "Invalid ID
format" error; specifically call parseMatrix(md) where md contains the `1.2` row
and assert parsed.errors.some(e => e.includes('Invalid ID format')) is true (and
optionally assert the malformed row is not present in parsed.rows). Use the
existing test style and functions (parseMatrix, parsed.errors, parsed.rows) to
locate where to add this new case.
🪄 Autofix (Beta)
Fix all unresolved CodeRabbit comments on this PR:
- Push a commit to this branch (recommended)
- Create a new PR with the fixes
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: 4d561acf-ddb9-4a4f-8b73-59e668a2891c
📒 Files selected for processing (9)
.github/PULL_REQUEST_TEMPLATE.md.github/workflows/pr-quality.ymlapp/test/checklist-parser.test.tsapp/test/coverage-matrix-parser.test.tsscripts/check-coverage-matrix.mjsscripts/check-pr-checklist.mjsscripts/feature-ids.jsonscripts/lib/checklist-parser.mjsscripts/lib/coverage-matrix-parser.mjs
✅ Files skipped from review due to trivial changes (3)
- scripts/feature-ids.json
- .github/workflows/pr-quality.yml
- .github/PULL_REQUEST_TEMPLATE.md
🚧 Files skipped from review as they are similar to previous changes (3)
- scripts/check-pr-checklist.mjs
- scripts/check-coverage-matrix.mjs
- scripts/lib/checklist-parser.mjs
Comment on lines
+1
to
+16
| const ROW_REGEX = /^\| (\d+(?:\.\d+){2,3}) \| ([^|]+) \| ([^|]+) \| ([^|]+) \| ([^|]+) \| ([^|]+) \|/; | ||
| const ID_REGEX = /^\d+(?:\.\d+){2,3}$/; | ||
| const VALID_STATUS = new Set(['✅', '🟡', '❌', '🚫']); | ||
|
|
||
| export function parseMatrix(markdown) { | ||
| const rows = []; | ||
| const errors = []; | ||
| if (typeof markdown !== 'string') { | ||
| return { rows, errors }; | ||
| } | ||
| for (const line of markdown.split(/\r?\n/)) { | ||
| const match = line.match(ROW_REGEX); | ||
| if (!match) continue; | ||
| const [, id, name, layer, path, status, notes] = match.map((v) => (typeof v === 'string' ? v.trim() : v)); | ||
| if (!ID_REGEX.test(id)) { | ||
| errors.push(`Invalid ID format: ${id}`); |
Contributor
There was a problem hiding this comment.
Malformed IDs are silently dropped instead of reported.
Because Line 1 already requires a valid ID shape, malformed IDs never reach the Line 15 validation branch. That bypasses Invalid ID format errors and weakens the gate.
Proposed fix
-const ROW_REGEX = /^\| (\d+(?:\.\d+){2,3}) \| ([^|]+) \| ([^|]+) \| ([^|]+) \| ([^|]+) \| ([^|]+) \|/;
+const ROW_REGEX = /^\|\s*([0-9][^|]*)\s*\|\s*([^|]+)\s*\|\s*([^|]+)\s*\|\s*([^|]+)\s*\|\s*([^|]+)\s*\|\s*([^|]*)\s*\|$/;
const ID_REGEX = /^\d+(?:\.\d+){2,3}$/;📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| const ROW_REGEX = /^\| (\d+(?:\.\d+){2,3}) \| ([^|]+) \| ([^|]+) \| ([^|]+) \| ([^|]+) \| ([^|]+) \|/; | |
| const ID_REGEX = /^\d+(?:\.\d+){2,3}$/; | |
| const VALID_STATUS = new Set(['✅', '🟡', '❌', '🚫']); | |
| export function parseMatrix(markdown) { | |
| const rows = []; | |
| const errors = []; | |
| if (typeof markdown !== 'string') { | |
| return { rows, errors }; | |
| } | |
| for (const line of markdown.split(/\r?\n/)) { | |
| const match = line.match(ROW_REGEX); | |
| if (!match) continue; | |
| const [, id, name, layer, path, status, notes] = match.map((v) => (typeof v === 'string' ? v.trim() : v)); | |
| if (!ID_REGEX.test(id)) { | |
| errors.push(`Invalid ID format: ${id}`); | |
| const ROW_REGEX = /^\|\s*([0-9][^|]*)\s*\|\s*([^|]+)\s*\|\s*([^|]+)\s*\|\s*([^|]+)\s*\|\s*([^|]+)\s*\|\s*([^|]*)\s*\|$/; | |
| const ID_REGEX = /^\d+(?:\.\d+){2,3}$/; | |
| const VALID_STATUS = new Set(['✅', '🟡', '❌', '🚫']); | |
| export function parseMatrix(markdown) { | |
| const rows = []; | |
| const errors = []; | |
| if (typeof markdown !== 'string') { | |
| return { rows, errors }; | |
| } | |
| for (const line of markdown.split(/\r?\n/)) { | |
| const match = line.match(ROW_REGEX); | |
| if (!match) continue; | |
| const [, id, name, layer, path, status, notes] = match.map((v) => (typeof v === 'string' ? v.trim() : v)); | |
| if (!ID_REGEX.test(id)) { | |
| errors.push(`Invalid ID format: ${id}`); |
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@scripts/lib/coverage-matrix-parser.mjs` around lines 1 - 16, ROW_REGEX is too
strict so malformed IDs don't match the row pattern and never hit the ID_REGEX
validation in parseMatrix; change ROW_REGEX (used by parseMatrix) to capture the
first table cell generically (e.g. accept any non-pipe content) instead of
enforcing the dotted-number shape, then keep the existing ID_REGEX check to push
`Invalid ID format: ${id}` into errors when the captured id fails validation;
update the ROW_REGEX constant and ensure parseMatrix still trims and validates
the captured id against ID_REGEX and records errors for bad IDs.
…i#965) The brew tap repo doesn't exist yet (planned, not created). Pre-existing in docs/install.md. Add exclude so the gate flags only new dead links. Refs tinyhumansai#965
senamakel
approved these changes
Apr 29, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Problem
#773 / PR #980 landed the test coverage matrix and testing strategy but left two contracts unenforced:
Plus markdown link rot: the strategy doc references `docs/RELEASE-MANUAL-SMOKE.md` (#971) and the new domain READMEs (#966), and we had no link-check job.
Solution
PR template (`.github/PULL_REQUEST_TEMPLATE.md`)
6 schema-locked items with explicit N/A guidance: tests added, matrix updated, feature IDs listed, no new external network deps, manual smoke updated if release-cut surface, `Closes #NNN` keyword used.
CI workflow (`.github/workflows/pr-quality.yml`)
Three jobs running on `ubuntu-latest`, all `continue-on-error: true`:
All three skip when the PR has the `docs` or `chore` label so trivial doc edits don't block.
Pure parsers + CLI shims
Node stdlib only — no new `package.json` dependencies.
Submission Checklist
Impact
Related
Summary by CodeRabbit
Chores
New Features
Tests