Standards Index Skill
Keep all standard documents indexed and discoverable by maintaining category README files.
Context
Every standard document must appear in the appropriate category README, or it becomes "orphaned" — undiscovered and unused. This skill teaches you to scan, verify, and fix the standards index structure.
Domain Context
Standards are organized in docs/standards/ by category:
engineering/— Code patterns, architecture, APIsdesign/— UI, styling, accessibilityquality/— Testing, governance, documentationoperations/— Deployment, monitoring, incidentsarchitecture/— Module boundaries, package design
Each category has a README.md that indexes all documents in that category. The index is the discoverable interface to all standards.
Reference: levels/L1-context/library/ structure
Instructions
Scan all category directories — Find every
.mdfile that is not namedREADME.mdfind docs/standards -type f -name "*.md" ! -name "README.md" | sort- Deliverable: Complete list of all standard files
Verify each file in the README — Check that each
.mdappears in its category'sREADME.md- For each file found in step 1, search for it in the corresponding
README.md - Format to look for:
- [Title](./filename.md) — Description - Deliverable: List of files that ARE indexed (matches) and files that are NOT (orphaned)
- For each file found in step 1, search for it in the corresponding
Run check-doc-index.sh — Use verification script to catch any misses
./levels/L2-ceremonies/scripts/check-doc-index.sh- This script identifies orphaned documents and invalid references
- Deliverable: Script output showing status
Fix orphaned documents — For each orphaned file:
- Option A: Add reference to the appropriate category README
- Option B: Delete the file if it's not needed (discuss with team first)
- Option C: Move the file to the correct category
- Update the README with alphabetized entry:
- [Standard Name](./filename.md) — One sentence description - Deliverable: All orphaned docs either indexed or removed
Verify index coherence — Check that category READMEs are well-formed
- Each README has a brief intro explaining the category
- All entries are alphabetized by title
- All descriptions are 1-line, scannable, consistent in format
- Deliverable: Coherent, well-organized category indexes
Anti-Patterns
Orphaned standards — Documents that exist but are not indexed
- Happens when: A standard is created but not added to README
- Impact: Standard is never discovered or used
- Guard: Before creating any
.mdfile indocs/standards/, plan its README entry
Stale index entries — README links to files that no longer exist
- Happens when: A file is deleted or moved, but the README is not updated
- Impact: Broken references in agent context
- Guard: When deleting or moving a standard, immediately update the README
Inconsistent formatting — README entries don't follow a pattern
- Happens when: Multiple people edit READMEs without a template
- Impact: Index is hard to scan and maintain
- Guard: Use the template format consistently:
- [Title](./file.md) — Description
Further Reading
levels/L2-ceremonies/scripts/check-doc-index.sh— Verification scriptlevels/L1-context/library/README.md— Example of well-maintained indexAGENTS.md— How agent files reference the standards index