duyetbot-knowledge

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

• Third-party content exposure detected (high risk: 0.90). The SKILL.md explicitly instructs the agent to fetch and ingest public, user-generated sources (e.g., https://duyet.net/llms.txt, https://blog.duyet.net/feed, https://github.com/duyet, and https://x.com/_duyet) as part of required update workflows, so untrusted third-party content could directly influence the agent's knowledge and actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly fetches external "LLM-friendly" files at runtime (see the quick-update step ./scripts/fetch-duyet-data.sh and the Knowledge Sources) — specifically https://duyet.net/llms.txt, https://cv.duyet.net/llms.txt, and https://blog.duyet.net/llms.txt — which are intended to be injected into the agent's knowledge/context and therefore directly control prompts.