backend-dev

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

• Third-party content exposure detected (high risk: 0.85). SKILL.md:7-10 describes fetching/caching documentation content (including via runtime “WebFetch URL” like https://jeesite.com/docs/...); that fetched outsider-authored web page text is then read from cached markdown files and fed into the agent’s LLM context.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's runtime cache step runs a fetch script (python3 scripts/cache_docs.py ...) that pulls documentation from external URLs such as https://jeesite.com/docs/install-deploy/ and then reads that fetched content to form responses, so remote content directly controls the agent's prompts.