# AgentTrust — Pre-Invocation AI Agent Trust & Reliability Checks

> AgentTrust — Pre-Invocation AI Agent Trust & Reliability Checks is a free MCP server for security. Check an AI agent endpoint's reliability and get a trust decision before calling it. It runs as a hosted remote server, so there is nothing to install locally.

- **Kind:** MCP server
- **Pricing:** Free
- **Platforms:** API, CLI
- **Website:** https://getagenttrust.com/docs
- **Repository:** https://github.com/agenttrust-ai/agenttrust
- **MCP Registry name:** io.github.agenttrust-ai/agenttrust
- **Categories:** security, developer tools
- **Listing updated:** 2026-11-16

## About

AgentTrust lets one agent look up another before invoking it. Given the exact endpoint URL, it returns the reliability evidence already collected, endpoint-ownership verification and a machine-readable trustDecision, without ever contacting the endpoint being checked. The public connector needs no account or key and exposes two read-only tools, check_agent_trust and assess_agent.

Agents that register get health-checked once a day, can prove ownership of their endpoint and build a deterministic reliability score from observed history. A check on a known endpoint returns status, score and a decision with confidence and reason codes such as OWNERSHIP_VERIFIED or REACHABLE_AT_LAST_CHECK. assess_agent returns allow with caution, avoid or insufficient evidence, and an unknown URL yields UNKNOWN_ENDPOINT with no score.

The docs stress that the result is a signal from AgentTrust's own observations, not a certification or proof the agent is safe, so callers should keep their own safeguards. A TypeScript SDK needs Node 18 or later, and the REST API and keyed MCP endpoint use an API key. A monitored test agent is provided for trying it.

## Features

- **Pre-call trust check:** Looks up an endpoint before you invoke it.
- **Never contacts the target:** Uses stored observations only.
- **Ownership verification:** Reports whether the endpoint owner has proved control.
- **Reliability score:** Built from daily health checks.
- **Machine-readable decision:** Returns a decision, confidence and reason codes.
- **TypeScript SDK:** Typed helpers for checks and policy assertions.

## Good for

- Check an MCP server or agent URL before sending it work
- Block calls to endpoints with no evidence
- Add a trust policy to an orchestration script
- Register your own agent to build a reliability record

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http agenttrust https://getagenttrust.com/api/mcp/public
```

### Claude Desktop (Settings → Connectors → Add custom connector)

```sh
https://getagenttrust.com/api/mcp/public
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "agenttrust": {
      "url": "https://getagenttrust.com/api/mcp/public"
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "agenttrust": {
      "type": "http",
      "url": "https://getagenttrust.com/api/mcp/public"
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "agenttrust": {
      "serverUrl": "https://getagenttrust.com/api/mcp/public"
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "agenttrust": {
      "type": "http",
      "url": "https://getagenttrust.com/api/mcp/public"
    }
  }
}
```

## Tools

- `check_agent_trust`: Check an endpoint URL and return reliability evidence and a trust decision.
- `assess_agent`: Return a use-or-avoid assessment for an endpoint URL.

## FAQ

### Do I need an API key?

Not for the public MCP connector or anonymous SDK checks; the REST API needs a key.

### Does it call the endpoint I check?

No, it only returns data it has already collected.

### Does a good result prove safety?

No, it is a signal for you to weigh, not a certification.

### What if an endpoint is unknown?

The result is insufficient evidence with an UNKNOWN_ENDPOINT reason and no score.

### Is AgentTrust — Pre-Invocation AI Agent Trust & Reliability Checks free?

Yes. AgentTrust — Pre-Invocation AI Agent Trust & Reliability Checks is free to use.

### How do I add AgentTrust — Pre-Invocation AI Agent Trust & Reliability Checks to Claude Code?

Run this command in your terminal: claude mcp add --transport http agenttrust https://getagenttrust.com/api/mcp/public

### What are the best alternatives to AgentTrust — Pre-Invocation AI Agent Trust & Reliability Checks?

The closest alternatives to AgentTrust — Pre-Invocation AI Agent Trust & Reliability Checks are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

## Alternatives

- [REA](https://ai-powered.com/mcp/rea): REA is an open-source MCP server for developer tools.
- [Codebase Memory](https://ai-powered.com/mcp/codebase-memory): Codebase Memory is an open-source MCP server for knowledge and memory.
- [GitHub](https://ai-powered.com/mcp/github-2): GitHub is an open-source MCP server for developer tools.
- [Serena MCP: the IDE for your agent](https://ai-powered.com/mcp/serena-mcp-the-ide-for-your-agent): Serena MCP: the IDE for your agent is an open-source MCP server for developer tools.
- [XERJ](https://ai-powered.com/mcp/xerj): XERJ is an open-source MCP server for knowledge and memory.
- [Ouroboros](https://ai-powered.com/mcp/ouroboros): Ouroboros is an open-source MCP server for developer tools.
- [Claude Code Ultimate Guide](https://ai-powered.com/mcp/claude-code-ultimate-guide): Claude Code Ultimate Guide is an open-source MCP server for ai models and llm tooling.
- [Concord MCP](https://ai-powered.com/mcp/concord-mcp): Concord MCP is a freemium MCP server for developer tools.

---
Source: https://ai-powered.com/mcp/agenttrust-pre-invocation-ai-agent-trust-and-reliability-che
