Trust Cop: Building a Security Agent for MCP Tool Drift with TrueForge

说明与用途

Trust Cop 监控 MCP 工具定义漂移，并在代理使用更新后的工具前要求人工批准，防止隐藏的安全回退。

问题

An AI agent trusts an MCP tool based on its initial definition, but the tool’s definition can change silently after approval, potentially compromising security.

结果

Trust Cop ensures that any change to an MCP tool’s definition is detected, reviewed, and either approved or denied before the agent can use the tool, preventing silent security regressions.

工作流

• 1. Agent approves a tool and its definition is stored as a baseline.
• 2. Guardian periodically fingerprints the live tool definition and compares it to the baseline.
• 3. If differences are detected, Guardian flags the tool as blocked_pending_review.
• 4. Trust Cop retrieves pending drift events and presents changes in plain language to a human reviewer.
• 5. Human approves or denies the change via TrueForge’s approval workflow.
• 6. If approved, Guardian restores tool access; if denied, the tool remains blocked and the decision is logged.

证据

Guardian fingerprints that definition and keeps checking the live version against it on a regular cycle. If something changes, Guardian flags the difference and moves the tool into: blocked_pending_review

来源

Trust Cop explains what changed in plain terms and asks a human to make the call. That step is backed by TrueForge's built-in human approval mechanism

来源