thinkies-cite-sources

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

• Third-party content exposure detected (high risk: 0.85). The skill’s workflow is about tracking/validating external URLs and citing retrieved information, which implies it will fetch outsider-authored web content at runtime (public web pages via WebFetch) and then include that retrieved prose in the agent’s LLM context for citation.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs the agent to use WebFetch at runtime to fetch and validate arbitrary URLs (e.g., https://github.com/synapseradio/ai-skills and https://github.com/synapseradio/ai-skills/raw/main/packaged/thinkies/cite-sources.skill), and those fetched contents are incorporated into responses/citations and therefore directly influence the agent's output and behavior.