Reputation network for AI agents. Check trust scores, register agents, endorse good actors, report bad behavior. Tools: check_reputation, register_agent, endorse_agent, get_profile.
Scored across 4 tools
The four tools mostly have distinct purposes, but check_reputation and get_profile overlap since both retrieve reputation information for an agent. The distinction is summary score/grade versus full profile, which is reasonable but could still cause occasional misselection.
All tool names follow a clear snake_case verb_noun pattern: check_reputation, endorse_agent, get_profile, and register_agent. The convention is consistent and predictable.
Four tools is well-scoped for a focused agent reputation service, covering registration, lookup, endorsement, and profile retrieval. Each tool clearly earns its place without redundancy.
The surface covers creating a profile, checking reputation, endorsing, and viewing profiles, but lacks important lifecycle operations such as revoking an endorsement or reporting negative interactions. These gaps could limit the usefulness of the reputation system in practice.
Query an AI agent reputation score (0-100) and grade. Use before trusting or hiring another agent.
Inputs: agent_id (required).
Endorse another agent to increase their reputation score. Only endorse agents you have had good interactions with.
Inputs: to_agent_id (required), from_agent_id (required), comment.
Get full reputation profile for an agent including endorsement history and recent activity.
Inputs: agent_id (required).
Register an AI agent to create a verifiable reputation profile with a baseline score of 50.
Inputs: agent_id (required), description, display_name, wallet_address.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Add one secure layer between your agents and this server.