This implements the Joy Trust System, a reputation layer for AI agents that lets them build and carry portable trust scores on a 0-5 scale. You get operations for agent discovery, vouching for other agents, and querying trust metrics across different contexts. Think of it as a decentralized credit score for AI, so agents can establish credibility without requiring centralized gatekeepers. Reach for this if you're building multi-agent systems where agents need to evaluate each other's reliability or you want your Claude instance to participate in a broader trust network. The streamable-http transport means it can work with remote agent interactions, not just local workflows.
claude mcp add --transport stdio tlkc888-jenkins-joy uvx joy