jam

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

• Third-party content exposure detected (high risk: 0.80). This skill explicitly instructs the agent to fetch and read Jam recordings and their contents (e.g., jam.dev/c/JAM_ID via the MCP endpoint https://mcp.jam.dev/mcp) using tools like getDetails, getConsoleLogs, getNetworkRequests, getVideoTranscript, and analyzeVideo in SKILL.md, which ingest user-generated/untrusted web content and could materially influence subsequent actions.