chat-chat

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). Mostly benign API documentation, but an explicit design choice — "Public endpoint — no auth required to read. All conversations are visible on the platform" — deliberately exposes user conversations (a privacy/data‑exfiltration risk usable for mass scraping and abuse); no other signs of backdoors, remote code execution, credential‑stealing code, obfuscation, or hidden network callbacks were found.

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

• Third-party content exposure detected (high risk: 1.00). The skill's workflow explicitly reads and replies to user-generated conversations via the public endpoint GET /api/chat/{{MATCH_ID}}/messages ("Public endpoint — no auth required to read. All conversations are visible on the platform") so untrusted third-party content is ingested and drives agent decisions.