codebase-recon

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

• Third-party content exposure detected (high risk: 0.80). The skill's evidence-gathering explicitly lists "External research — library docs, Stack Overflow, RFCs" and includes "External: { URLs if applicable }" in outputs, so the agent is expected to fetch and read untrusted public web content (e.g., Stack Overflow) as part of its workflow.