audit-code-review

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.75). The workflow explicitly instructs using a runtime web research tool (CallMcpTool with server: "user-firecrawl" and sources: [{ "type": "web" }] in SKILL.md), which can fetch and ingest outsider-authored free-text web content into the agent’s LLM context.