Agent♥︎Age
Catalog

Routescore

Official8 toolsLive

by routescore · TypeScript

Read-only pre-sign evidence for onchain agents: route quality, MEV exposure, clear/caution verdict.

io.routescore/mcp — MCP Server for Read-Only Pre-Sign Evidence

The io.routescore/mcp server provides a read-only “pre-sign evidence” layer for onchain agents. It exposes check_swap to return a modeled read of route quality and MEV/execution exposure, plus token-registry recognition. Results are provided as a clear / caution / unsupported verdict with caveats, alongside a hash-verifiable pre-sign journal of what was known before signing.

🛠️ Key Features

  • Read-only pre-sign evidence (“pre-sign journal”)
  • check_swap returns route quality, MEV/execution exposure, and token-registry recognition
  • clear / caution / unsupported verdicts with caveats
  • Persists hash-verifiable records of pre-sign knowledge

🚀 Use Cases

  • Pre-sign review for onchain swap actions by agents or users
  • Offline/re-verifiable audit of what an agent knew before signing

⚡ Developer Benefits

  • Supports hash-verifiable, offline re-verification of pre-sign observations
  • Fits MCP-based workflows for onchain agents

⚠️ Limitations

  • Does not sign, execute, route funds, or custody assets
  • Produces modeled reads with verdict caveats (not execution guarantees)

📚 Topics

ai-agents, defi, ethereum, mcp, mev, model-context-protocol, onchain-agents, pre-sign, routescore, typescript

Topics

ai-agentsdefiethereummcpmevmodel-context-protocolonchain-agentspre-signroutescoretypescript

Related servers

More in Security