AgentTrust — Identity & Trust for A2A Agents

A
8.0

Identity, trust, and A2A orchestration for autonomous AI agents. Official A2A partner.

Installation

Install from source

git clone https://github.com/agenttrust/mcp-server
cd ai-agenttrust-mcp-server
npm install
npm start

Security Report

Score Breakdown

Description10
Permissions10
Behavior6
Stability--

Findings (5)

medium
vague-description

Unclear Server Purpose and Capabilities

The description uses broad, abstract language ('identity, trust, and A2A orchestration') without specifying concrete capabilities or what tools/resources this server actually provides. Terms like 'Official A2A partner' lack context and verification.

medium
excessive-scope

Broad Scope Without Clear Boundaries

The server claims to handle 'identity, trust, and A2A orchestration' - three distinct domains. Without tool definitions visible, it's unclear what the actual scope is and whether capabilities are proportional to the stated purpose.

low
info

Transport Type Unknown

The transport mechanism is listed as 'unknown'. This makes it difficult to assess exposure risk (local stdio vs. remote HTTP endpoint).

low
info

No Tools Available for Review

Tool definitions were not fetched from the server. This prevents detailed analysis of input validation, prompt injection risks, and actual capabilities.

info
vague-description

Semantic Analysis Summary

AgentTrust has positive trust signals (GitHub repository, website, claimed A2A partnership) but suffers from vague description and unclear scope. The absence of visible tools and undefined transport type prevent thorough security assessment. The broad claims about 'identity, trust, and orchestration' without concrete capability examples raise concerns about scope proportionality.

Last scanned 1mo ago

Details

Version
1.1.1
Transport
Unknown
Capabilities