Let your agent
handle it.

Pick an Expert. Say one sentence to your agent — Claude Code, Codex, Cursor, or any AI assistant.
MCP, A2A, or any agent protocol — it handles the rest.

Official skills are curated & reviewed · Community skills are clearly labeled

Expert pick this week

Secure your agents before they go to production →

A security audit found 30 community skills silently enrolling agents into crypto swarms. Install Agent Sentinel to set up permission boundaries and injection detection — before your agents go to production.

Expert Install

Say to your agent
Read https://clawhub.md/expert/agent-security-expert.md and set me up as Agent Sentinel

Talk and handle it

  • "Set up permission boundaries for my agent"
  • "Block my agent from running sudo or rm -rf"
  • "Scan this URL for indirect prompt injection"
  • "Build an MCP server with scoped read-only access"

What are you trying to do?