広告掲載
MCP.so
ログイン

AgentAvow Trust MCP Server

@AgentAvow

リポジトリホームページドキュメント

基本情報

カテゴリ

AI とエージェント

トランスポート

stdio

作者

AgentAvow

投稿者

kenneives

AgentAvow Trust MCP Server について

Trust verification and security scanning for AI agents. Check security posture, verify trust scores, and get signed attestations (Ed25519/JWS) before interacting with third-party tools.

#security#code-scanning#mcp#supply-chain#attestation

AgentAvow Trust MCP Server の設定

以下の設定を使って、このサーバーを MCP 対応クライアントに追加してください。

{
  "mcpServers": {
    "agentavow": {
      "type": "http",
      "url": "https://agentavow.com/mcp"
    }
  }
}

AgentAvow Trust MCP Server のツール

ツール一覧はまだありません

この掲載には表示できるツール一覧がまだありません。サーバーにツールがないという意味ではありません。対応機能はプロジェクトのドキュメントをご確認ください。

上のボタンでサーバーのエンドポイントに接続し、ツール一覧を取得できます。

概要

What is AgentAvow Trust?

AgentAvow Trust is an MCP server that grades the safety of anything an AI agent connects to — a GitHub repo, an MCP server, or an npm/PyPI/crates/Docker/Hugging Face package. It returns a signed, 0–100 trust score with a plain safe / needs-review verdict, backed by scan evidence anyone can recompute offline. Read-only and authless: no account, no keys.

How to use AgentAvow Trust

Add the remote server to your MCP client (Claude Code, Cursor, or your own runtime) — no install:

claude mcp add --transport http agentavow https://agentavow.com/mcp

Then ask it to check a tool before you connect: "scan the repo modelcontextprotocol/servers" or "scan the npm package chalk." Every result links to a full report and a signed attestation you can verify offline.

Key features

  • Signed, offline-recomputable verdicts — Ed25519/JWS over an RFC 8785 canonical verdict; recompute it byte-for-byte against the public JWKS, no callback required.
  • 12-category static analysis — secret hygiene, unsafe code execution, data handling, dependencies, prompt-injection surfaces, and more; every finding points at the exact file and line.
  • Scans across surfaces — GitHub repos, live MCP servers (it checks the tool definitions for poisoning/injection), and npm/PyPI/crates/Docker/Hugging Face packages.
  • Eight read-only tools — scan_repo, scan_package, scan_mcp_server, verify_trust, check_interaction_safety, lookup_identity, get_trust_badge, about_agentavow.
  • Authless — anonymous and read-only; nothing to configure.

Use cases

  • Scan an MCP server or package before your agent connects, so an unsafe tool is stopped in the loop rather than found after the fact.
  • Gate CI on a minimum score with the GitHub Action or CLI.
  • Check the interaction-safety threshold before delegating to or trading with another agent.
  • Add a signed trust badge to your README that anyone can verify.

FAQ

  • Do I need an account? No — anonymous and read-only.
  • Is it free? Yes, checking is free.
  • Can I verify a score without trusting AgentAvow? Yes — recompute the canonical verdict and check the Ed25519 signature offline against the public JWKS.

コメント

「AI とエージェント」の他のコンテンツ

L

Legion MCP

faulkj

MCP-native LLM councils for debates, juries, blind panels, voting, refinement, and custom multi-model deliberation.

1

bsv.cx

Andy Rowe

Timestamp and verify evidence on-chain. Let your agent prove what it saw and when. bsv.cx gives an AI agent verifiable memory. Timestamp a hash of any output, archive a web page as it looked, or fetch a URL with a recei

F

fhirHydrant

faulkj

Open-source Node.js FHIR MCP server with SMART Backend Services, metadata-aware search/CRUD tools, compact responses, FHIRPath filtering, safe pagination, audit events, and terminology lookup.

14

OpenZiti / LLM-Gateway

openziti

NetFoundry LLM Gateway is an OpenAI-compatible API proxy that routes requests across multiple LLM providers using zero-trust networking. Zero trust LLM gateway. OpenAI-compatible proxy with semantic routing and load bala

93

Krafters

Krafters (Romain Duclos)

Krafters is a French club of specialized AIs for freelancers and small businesses. Its connector opens the member's AI kits for Facebook and Instagram ads, SEO/GEO and short videos directly in Claude, and gives the next

広告掲載