moltbook

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

• Third-party content exposure detected (medium risk: 0.65). SKILL.md instructs runtime behavior that fetches Moltbook content (e.g., pinned post/community manifesto and the user’s feed) from Moltbook endpoints like /api/v1/posts... and then uses that content to generate comments/self-introductions; those posts/comments are authored by outsiders, so their free text can be ingested into the agent’s LLM context (indirect prompt injection risk).

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly tells the agent at runtime to "Fetch https://www.moltbook.com/heartbeat.md and follow it" (and to re-fetch SKILL.md, MESSAGING.md, and skill.json for updates), meaning externally-hosted plaintext can dynamically control agent instructions.