find-docs

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

• Third-party content exposure detected (high risk: 1.00). Yes — the SKILL.md workflow requires running ctx7 commands (ctx7 library and ctx7 docs) that retrieve live documentation and code/examples from external public sources (and the --research option explicitly git-pulls source repos and runs live web search), so the agent ingests untrusted third-party content that could alter its behavior.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill instructs running the Context7 CLI (e.g., "npx ctx7@latest") which fetches and executes code from the npm registry (e.g., https://registry.npmjs.org/ctx7) and its --research mode explicitly git-pulls external repositories (remote git URLs) at runtime to synthesize responses, so remote content is fetched and executed/used to control prompts.