Skip to content

feat(ci): add AI content quality guardrails - #68

Merged
oskarth merged 5 commits into
masterfrom
feat/ai-content-guardrails
Jan 26, 2026
Merged

feat(ci): add AI content quality guardrails#68
oskarth merged 5 commits into
masterfrom
feat/ai-content-guardrails

Conversation

@oskarth

@oskarth oskarth commented Jan 23, 2026

Copy link
Copy Markdown
Collaborator

Summary

  • Add Vale prose linter with IPTF custom styles (terminology, marketing language, hedging)
  • Add GLOSSARY.md term consistency checker
  • Extend validate-patterns.js to cover all content types with JSON Schema validation
  • Add Husky pre-commit hooks for automated checks
  • Add on-demand LLM review tool for pattern evaluation

Test plan

  • npm run validate passes
  • npm run check-terms passes
  • npm run lint:vale runs without errors (warnings expected for existing content)
  • Pre-commit hooks trigger on staged files
  • CI workflow added for GitHub Actions

Closes #58

oskarth and others added 3 commits January 23, 2026 12:10
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Implements Issue #58 improvements to guard against LLM hallucinations:

- Vale prose linter with IPTF styles (marketing, hedging, terminology)
- GLOSSARY.md term consistency checker (scripts/check-terminology.js)
- Extended validation for all content types (vendors, use-cases, approaches, jurisdictions)
- JSON Schema validation for frontmatter
- Husky pre-commit hooks with lint-staged
- On-demand LLM review tool using Claude API (scripts/llm-review.js)
Delete the empty Vocab.yml rule file that was triggering 498 false
positives. Vocabulary handling is already correct via accept.txt.

Also add Development section to README with validation commands.
@oskarth
oskarth merged commit e3245b0 into master Jan 26, 2026
5 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

CI/Linting: Future improvements for AI-generated content quality

2 participants