consolidate-research

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly ingests public web content: the manifest/schema lists models[].capabilities including web_search and site_restrictions (references/consolidation-manifest-schema.md), the normalization step extracts "web search findings with source URLs" (SKILL.md Step 2), and the Agentic mode and Reconcile steps perform web searches to resolve conflicts and drive decisions—so untrusted third‑party pages (e.g., GitHub, docs, HN) are fetched and used to influence actions.