Low Risk — Risk Score 15/100
Last scan:2 days ago Rescan
15 /100
agent-vitamins
Daily self-improvement for AI agents. Fetches curated briefs from Agent Vitamins with actionable improvements.
A legitimate documentation-only skill that fetches curated AI agent improvement briefs via MCP tools. No code execution, filesystem access, or credential handling. External service reference is clearly disclosed.
Skill Nameagent-vitamins
Duration20.3s
Enginepi
Safe to install
Safe to use. The skill is well-documented and requires user approval before any recommended improvements are implemented.
ResourceDeclaredInferredStatusEvidence
Filesystem NONE NONE No filesystem access declared or inferred
Network NONE NONE Network access delegated to MCP server; skill only provides recommendations
Shell NONE NONE No shell execution in skill
Environment NONE NONE No environment variable access
Skill Invoke NONE NONE Skill only makes recommendations, does not invoke other skills
Clipboard NONE NONE No clipboard access
Browser NONE NONE No browser automation
Database NONE NONE No database access
1 findings
🔗
Medium External URL 外部 URL
https://agentvitamins.com
SKILL.md:32

File Tree

1 files · 5.1 KB · 138 lines
Markdown 1f · 138L
└─ 📝 SKILL.md Markdown 138L · 5.1 KB

Dependencies 1 items

PackageVersionSourceKnown VulnsNotes
mcp-agent-vitamins not specified npm (npx) No External MCP package; verify integrity before use

Security Positives

✓ Clear documentation with well-defined behavior and rules
✓ User approval required before implementing any recommendations
✓ No automatic code execution - purely advisory role
✓ No filesystem, shell, or credential access
✓ Transparent about external service dependency (agentvitamins.com)
✓ Honest skipping policy documented - explains why items are skipped