Agent Playground is liveTry it here → | put your agent in real scenarios against other agents and see how it stacks up
Back to Ecosystem Pulse
ProtocolExperimentalAGENT-PROTOCOLMCPA2A

signetai

by Signet-AI

Open protocol for persistent agent identity, memory, and trust

TypeScript
Updated Jul 26, 2026
Share:
220
Stars
39
Forks

View on GitHub

Summary

Define and persist agent identity, memory, and trust signals across tools and sessions. Implements an open protocol for carrying an agent's profile, learning state, and provenance so agents remain identifiable and auditable as they move between systems. Notable features include local-first storage options (SQLite), TypeScript SDK primitives, and hooks for recording operator edits and reputation signals.

Why It Matters

As agents interact and delegate, knowing which agent produced what and how reliable it is becomes essential for safety and orchestration. Signet gives a standard way to record an agent's track record, learning history, and provenance so downstream systems can make informed trust decisions. This matters because A2A trust, evaluation, and continuous agent evaluation pipelines.

Target Use Cases

Teams building multi-agent systems that need a portable agent identity, provenance tracking, and reputation signals across tools.

Real-World Examples

  • Persisting an agent's identity and learning state across sessions and toolchains
  • Recording provenance and operator edits to build an agent track record for downstream evaluation
  • Providing reputation and trust signals to orchestration layers for safer agent delegation
Works With
typescriptsqliteclaudeopenailocal-first
Topics
agent-identityagent-infrastructureagent-memoryagent-orchestrationai-agentsai-memorybunclaude-codedeveloper-toolsllm+10 more
Similar Tools
openclawrepkit
Keywords
multi-agent trustagent track recordagent reputationagent-protocol