Config auditor
Skill scoobydont-666/shared-claude-skills/skills/config-auditor
Cross-reference all CLAUDE.md files, memory files, skill descriptions, and project registries for contradictions, stale data, and inconsistencies. Trigger on "check for contradictions", "audit configs", "config consistency", "are my configs consistent", or any request to verify configuration coherence across Project Swarm.From its SKILL.md
npx -y skills add scoobydont-666/shared-claude-skills --skill config-auditorAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
4 things to look at
- skips confirmationTells the agent to proceed without asking first, 1 time: "Fix all non-ambiguous contradictions automatically.".
- 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
- runs commandsInstructs the agent to run 8 commands, including `ssh -o ConnectTimeout=5 <host> "nvidia-smi --query-gpu=name,memory.total --format=csv,noheader"` and 7 more.
- fetches URLsInstructs the agent to fetch 2 URLs, including http://127.0.0.1:<chromadb-port>/api/v2/collections and 1 more.
SKILL.md
3.0 KB, 711 tokens by cl100k_base, as published. Nobody here has run it
Config Auditor — Cross-Configuration Consistency Checker
When to Trigger
- "check for contradictions" / "audit configs" / "config consistency"
- After major infrastructure changes (hardware swaps, new hosts, repo changes)
- During skill-updater runs (call this as a sub-check)
- When config was updated in one place but might be stale elsewhere
What to Check
1. Hardware Specs (VERIFY FROM MACHINE)
# ALWAYS SSH and check before asserting specs
ssh -o ConnectTimeout=5 <host> "nvidia-smi --query-gpu=name,memory.total --format=csv,noheader"
ssh -o ConnectTimeout=5 <host> "lscpu | grep 'Model name\|Socket\|Core'"
ssh -o ConnectTimeout=5 <host> "free -h | head -2"
Cross-reference against:
~/.claude/CLAUDE.md(Cluster Infrastructure section)~/.claude/projects/*/memory/user_admin_user.md~/.claude/skills/ai-cluster-knowledge/SKILL.mddescription- Each host's memory file (project_<hostname>_server.md, etc.)
2. Repo Inventory
gh repo list your-github-user --limit 50 --json name,visibility,isPrivate
Cross-reference against:
project_repo_status.mdmemoryproject-management/SKILL.mdRepos sectionCLAUDE.mdGitHub section (visibility claims)
3. Skill Count
ls ~/.claude/skills/ | wc -l
Cross-reference against:
project_skill_updates.mdmemory
4. Port Assignments
Grep all CLAUDE.md files for port numbers, check for conflicts:
grep -rn "port [0-9]\|:[0-9][0-9][0-9][0-9]" /opt/*/CLAUDE.md ~/.claude/CLAUDE.md
5. ChromaDB Collection Prefixes
Cross-reference CLAUDE.md claims against actual collections:
# On GPU host:
curl -s http://127.0.0.1:<chromadb-port>/api/v2/collections | python3 -c 'import json,sys; ...'
6. NVIDIA Driver Version
ssh <host> "nvidia-smi --query-gpu=driver_version --format=csv,noheader"
vs all files that mention driver version.
7. Service Versions (Monero)
monerod --version
vs memory files.
8. Model Inventory
curl -s http://127.0.0.1:<ollama-port>/api/tags
vs model evaluation doc, ai-cluster-knowledge skill.
9. Project Phase Status
Cross-reference:
project-management/SKILL.mdProject Registry table- Each project's CLAUDE.md
- Each project's plans/*.md
Output Format
Table of contradictions found:
| # | Issue | File A says | File B says | Fix |
Fix all non-ambiguous contradictions automatically. Ask about ambiguous ones.
Key Rule
NEVER update hardware specs without SSHing to the machine first.
What ships with it
Read from the repository
Just SKILL.md. No reference files, no scripts.