mcp-integration-reference
Pass
Audited by Gen Agent Trust Hub on Apr 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a robust pattern for MCP integration by using reference documentation to guide the agent's tool usage, which helps prevent syntax errors and incorrect parameter usage.
- [PROMPT_INJECTION]: Indirect Prompt Injection Surface. The skill is designed to ingest and process external data (error logs and stack traces) from an observability platform (Sentry). * Ingestion points: Tool outputs from Sentry MCP tools (list_issues, get_issue, search_events). * Boundary markers: Absent; the instructions do not include explicit boundary markers or instructions to ignore potential directives embedded in the retrieved error data. * Capability inventory: The skill's capabilities are restricted to reading data via MCP tools and generating markdown reports. It does not utilize dangerous functions like eval, exec, or direct file system writes. * Sanitization: Absent; there is no explicit sanitization or filtering of the retrieved data before it is incorporated into the generated report. * Conclusion: While a surface for indirect prompt injection exists, the lack of high-privilege capabilities makes this a low-risk configuration.
Audit Metadata