ai-compassCAUTION
A plain-English map of AI models, MCP servers and agent skills, kept fresh by an agent
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
A plain-English map of AI models, MCP servers and agent skills, kept fresh by an agent.
Live: https://whichaimap.com/
- Repo Explainer: ruflo, OmniRoute, LiteLLM, Ollama... each popular AI repo in plain English (what, when, caveats, alternatives). Ask for any repo via an issue
- Model picker + benchmarks: tell it the job, get the model; heatmap of 9 benchmarks, per-skill rankings, smarts-vs-price chart (
data/benchmarks.json, every number linked to its source) - MCP guide + config generator: exact setup for Claude Code, Claude Desktop, claude.ai, ChatGPT, Codex, Cursor, VS Code, Gemini CLI
- Skills: what
SKILL.mdis, how to install, a curated list - GitHub 101: read a repo, judge if it's safe, install from it
- News: official blogs, agent releases, Hacker News, summarized by AI (Gemini or Claude)
- Leaderboard: vote for tools with 👍 on their GitHub issue, earn points
How it stays fresh
.github/workflows/agent.yml runs every 6 hours:
AI steps (news summaries, repo explanations, model watch) need one repo secret: GEMINI_API_KEY (used first) or ANTHROPIC_API_KEY. See agent/llm.mjs. Without a key, everything else still runs.
Points
+1 per 👍 vote · +2 per repo explained on request · +10 per accepted tool suggestion (issue labeled submission + accepted) · +5 per merged PR.
Edi
24412024cf06OBSERVED · 2026-10-10Host compatibility
What the documentation claims. We have not run a compatibility test.
| Host | Status | Notes |
|---|---|---|
| claude-code | mentioned | |
| claude-desktop | mentioned | |
| codex | mentioned | |
| copilot | mentioned | |
| cursor | mentioned | |
| gemini-cli | mentioned |
Trust audit
CAUTIONgrade B · trust 87/100 Install with care. The audit found things worth knowing before you trust its output.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | WARN |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | WARN |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- none-observed
- Network
- declared (7 observation(s))
- Shell
- none-observed
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (6)
api_url = "http://127.0.0.1"
# rp_origins = ["http://127.0.0.1:3000"]
@anthropic-ai/sdk, @google/genai, @supabase/supabase-js, zod
<ul class="vs-list no"><li>Read and change files anywhere you can, and start programs</li><li>Read your environment variables and settings, including API keys</li><li>See every prompt you send and eve
Gates applied: no_behavioural_pass, no_license.
24412024cf06full audit observations/trust-audit/skill/stefankasamakov__ai-compass.json · Report an issue / request a re-scanAudit history
Every audit this skill has had.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-10 | 24412024cf06 | CAUTION | B | 87 | first audit |
Questions
What does the ai-compass skill do?
A plain-English map of AI models, MCP servers and agent skills, kept fresh by an agent
Is ai-compass safe to install?
With care. The audit graded it B (87/100) and found 6 things worth knowing before you trust this skill, listed below with the exact line each was found on.
What can ai-compass access on my machine?
The audit observed that it reaches the network. Each of those is consistent with what it says it does. Secrets in the source: none found.
Which assistants does ai-compass work with?
Its documentation mentions claude-code, claude-desktop, codex, copilot, cursor and gemini-cli. That is what the text claims, not a compatibility test we ran.
How current is this page?
The grade is for one exact copy of the source (24412024cf06), read on 2026-10-10. The repository is watched, and a new audit runs when it changes — this is the first audit.