researching-with-deepwiki

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

• Third-party content exposure detected (high risk: 0.90). The skill routes analysis queries to the DeepWiki MCP server to fetch and analyze public repositories on GitHub, GitLab, and Bitbucket (e.g., github.com/owner/repo), which ingests untrusted, user-generated third‑party content that the agent reads and interprets for Q&A.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill is configured to call the DeepWiki MCP endpoint (https://mcp.deepwiki.com/mcp and the private variant https://mcp.devin.ai/deepwiki/mcp) at runtime to obtain analysis/instructions that directly drive the agent's behavior, making this an external runtime dependency that controls prompts.