Project Health

Project Health Check Skill

Internal skill used by the session-daily command. Checks the health of the AI DevStudio project itself.

Checks

1. Skill Count Consistency

Verify that the skill count (99) is consistent across all documentation files:

Bash: ls skills/ | wc -l

Then check these files reference the correct count:

• CLAUDE.md: grep for skill count
• AGENTS.md: grep for skill count
• README.md: grep for skill count

Report any inconsistencies found.

2. Stale Session Files

Bash: find .claude/sessions/ -name "*.json" -mtime +30 2>/dev/null | wc -l

Report count of session files older than 30 days. Suggest running /cleanproject if count > 5.

3. Cache Size

Bash: du -sh .claude/cache/ 2>/dev/null || echo "No cache directory"

Report cache size. Suggest cleanup if > 50MB.

4. Install Script Sync

Bash: python3 -c "import ast, sys; ast.parse(open('install.py').read()); print('install.py: OK')" 2>&1
Bash: bash -n install.sh 2>&1 && echo "install.sh: OK"

Report any syntax errors.

5. Documentation Currency

Check that docs/ modification dates are recent relative to skills/ changes:

Bash: git log --oneline -1 -- skills/ docs/

Output Format

Project Health: 

Skill count: N (consistent: yes/no)
Stale sessions: N files
Cache: 
Install scripts: OK / 
Docs currency: OK / 

Token Optimization

Expected range: 200-500 tokens

Patterns used: Bash-based queries, early exit if all checks pass

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: manastalukdar
• Source: manastalukdar/ai-devstudio
• License: MIT

Install and usage instructions live in the source repository linked above.