Skip to content
Squatch MCP

Agent Reliability

dev.agentreliability/agent-reliability

by dev.agentreliability· v0.6.14Source: Official MCP Registry

Findings

Detected by Squatch rules from the server's metadata and protocol behavior. Latest completed scan: .

Severity: High2 findings

Severity: HighStatus: OpenSQ-RESOURCE-INJECTION

Resource metadata contains directives to the AI model

A resource's metadata contains instructions addressed to the AI model.

First seen · Last seen

Suggested fix: Describe resources neutrally.

Severity: HighStatus: OpenSQ-RESOURCE-INJECTION

Resource metadata contains directives to the AI model

A resource's metadata contains instructions addressed to the AI model.

First seen · Last seen

Suggested fix: Describe resources neutrally.

What this scan does not cover

  • Squatch never calls a server's tools, so it cannot see what they actually do at run time.
  • It scans without credentials: tools behind authentication may not be visible.
  • It does not review source code, dependencies or the publisher's infrastructure.
  • A server can change behavior between scans; monitoring compares each scan with the last.
  • Classifications come from tool names, descriptions and schemas, which the publisher controls.

How Squatch scans servers