Strengthen portfolio evidence links and publication controls - #11
Conversation
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: dd9b6e6d3b
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| const findings = []; | ||
| for (const [index, line] of contents.split(/\r?\n/u).entries()) { | ||
| for (const rule of preparationRules) { | ||
| if (rule.expression.test(line)) findings.push({ line: index + 1, rule: rule.id }); |
There was a problem hiding this comment.
When a prohibited phrase is soft-wrapped across lines, as commonly happens in Markdown (for example, interview\npreparation), splitting the contents before applying rules means neither invocation sees the complete phrase; scanText therefore returns no findings and the repository publication check passes despite the rendered text containing the prohibited guidance. Scan the complete contents, or include adjacent-line overlap while retaining line-number calculation.
Useful? React with 👍 / 👎.
This branch was successfully deployed
1 participant
Motivation
Description
portfolio/docs/alignment-notes.mdand removed direct CTAs to unverified Document Intelligence docs fromportfolio/data/projects.tswhile retaining the project repo link and an accurate implementation description.portfolio/scripts/publication-content-check.mjs, a repository-root publication scanner that detects focused preparation/coaching language and suspicious document names, can optionally scan generated.nextoutput, and reports onlypath:line [rule]before returning a nonzero exit on findings.check:publicationandcheck:publication:buildtoportfolio/package.json, updating.github/workflows/portfolio-quality.ymlto run source checks before lint/build and generated-output checks after build, and documenting the checks inportfolio/README.md.portfolio/tests/content-contract.test.mjsto verify the new rules detect preparation guidance and avoid false positives on legitimate professional/technical language.Testing
npm run check:publication— passed locally with the new scanner reporting no failing findings.npm test— passed (10 tests).npm run lintandnpm run type-check— passed.npm run build— passed and produced static assets successfully.npm run check:publication:build— passed against generated output.npm run test:e2e— passed (2 production-server smoke tests covering anchors, details/accordion behavior, links-only contact policy, canonical metadata,robots.txt,sitemap.xml, and the generated social image).chore/portfolio-publication-standardsand the working tree contains only the intended edits; remote PR publication was prepared via the tooling but live remote push/merge is outside this task.Codex Task