deepwiki-hardened

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly connects to the DeepWiki MCP service (e.g., SKILL.md Base Server https://mcp.deepwiki.com/mcp and scripts/deepwiki.js using https://mcp.deepwiki.com/sse) to fetch public GitHub repository wikis and documentation (untrusted, user-generated content) which the agent reads and uses to generate answers, so that third‑party content could carry instructions that materially influence the agent's behavior.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill's runtime script establishes an SSE connection to https://mcp.deepwiki.com/sse (and posts to endpoints under https://mcp.deepwiki.com) to fetch remote content that is injected into the agent's output/response, making that host a required runtime dependency which can directly control the agent's presented prompts/content.