yao-doctor-skill

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.85). The list mixes trusted references (official GitHub projects, arXiv, api.github.com, X) with multiple high-risk items — direct links to .sh/.py files, paste/raw hosts (paste.rs), and Slack webhook endpoints — which are common vectors for distributing or exfiltrating malware and therefore make the set suspicious.

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

• Third-party content exposure detected (high risk: 0.75). Runtime path: run_yao_doctor_skill.py scans discovered skill/workbench directories (including outsider-authored SKILL.md/code text) and then renders those findings into report.html/report.json; if the report is later used as LLM context, that outsider free text can be ingested (indirect prompt injection risk).