Agentโ™ฅ๏ธŽAge
Catalog

io.github.lucaslubi/fidacy-mcp

Official

by fidacy ยท TypeScript

Action firewall for AI agents: blocks the wrong action before it runs, every verdict Ed25519-signed.

An MCP server implementing an action firewall for AI agents. It blocks the wrong action before it runs, using an integrity mechanism where every verdict is Ed25519-signed. The server is identified by the slug io-github-lucaslubi-fidacy-mcp and exposes the behavior described in its name and description.

๐Ÿ› ๏ธ Key Features

  • Action firewall behavior for AI agents
  • Blocks incorrect actions before execution
  • Ed25519-signed verdicts for each decision

๐Ÿš€ Use Cases

  • Enforcing safety by preventing disallowed or incorrect agent actions prior to running them
  • Verifying firewall verdicts with Ed25519 signatures

โšก Developer Benefits

  • Signed verdicts can support auditability and integrity checks
  • Clear pre-execution gating for agent actions
io.github.lucaslubi/fidacy-mcp - agentage MCP Catalog