oathe

oathe / scan / reports

Audited agent skills and MCP servers

Every skill here was installed in an isolated sandbox and watched while it ran — file access, network calls, process activity. The score reflects what it did, not what its README claims. 2,986 skills audited so far.

Browse by verdict: malicious · dangerous · caution · safe · audit a new skill

oathe-ai/darwin 67 CAUTION Full report →
oathe-ai/oathfish 77 CAUTION Full report →
oathe-ai/linkedin-growth-engine 74 CAUTION Full report →
intuit/quickbooks-online-mcp-server 95 SAFE Full report →
oathe-ai/oathe-research 76 CAUTION Full report →
skills/git-pushing 60 CAUTION Full report →
codexstar69/bug-hunter 80 SAFE Full report →
ship-learn-next 91 SAFE Full report →
productivity/grill-me 96 SAFE Full report →
skills/browser-use 72 CAUTION Full report →
skills/brainstorming 86 SAFE Full report →
skills/ui-ux-pro-max 88 SAFE Full report →
skills/frontend-design 97 SAFE Full report →
juliusbrussee/caveman 76 CAUTION Full report →
skills/immigration-document-checklist 98 SAFE Full report →
skills/legal-writer 82 SAFE Full report →
skills/signature-request 95 SAFE Full report →
skills/legal-risk-assessment 97 SAFE Full report →
skills/legal-response 92 SAFE Full report →
skills/brief 96 SAFE Full report →
skills/contract-and-proposal-writer 95 SAFE Full report →
skills/general-counsel-advisor 84 SAFE Full report →
tommyjepsen/awesome-ux-skills 90 SAFE Full report →
muratcankoylan/agent-skills-for-context-engineering 88 SAFE Full report →
Graphify-Labs/graphify 68 CAUTION Full report →
Fandhe-AI/agent-cli-skills 86 SAFE Full report →
AikidoSec/aikido-claude-plugin 70 CAUTION Full report →
usestrix/strix 77 CAUTION Full report →
Fission-AI/OpenSpec 83 SAFE Full report →
addyosmani/agent-skills 92 SAFE Full report →
cloudflare/computer 93 SAFE Full report →
tanweai/pua 68 CAUTION Full report →
diegosouzapw/OmniRoute 49 DANGEROUS Full report →
nextlevelbuilder/ui-ux-pro-max-skill 76 CAUTION Full report →
IlhamKassim/resume-builder 88 SAFE Full report →
.skills/media 91 SAFE Full report →
composio-community/skills 66 CAUTION Full report →
ZeroPointRepo/youtube-skills 87 SAFE Full report →
search/scielo-api 97 SAFE Full report →
design-templates/web-prototype-taste-editorial 97 SAFE Full report →
Forward-Future/loopy 87 SAFE Full report →
AMAP-ML/SkillClaw 71 CAUTION Full report →
kepano/defuddle 96 SAFE Full report →
AviVAvi/TokenScope 78 CAUTION Full report →
eph5xx/tiebreaker 93 SAFE Full report →
clawhub/hightower6eu--google-workspace-2z5dp 19 MALICIOUS Full report →
jasonxkensei/xProof 67 CAUTION Full report →
coreyhaines31/marketingskills 92 SAFE Full report →
davidondrej/skills 82 SAFE Full report →
Panniantong/agent-reach 76 CAUTION Full report →

Scores run 0–100. Every audit runs the skill in a sandbox and records what it actually does. How the audit works · What the verdicts mean