vibehub

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

• Third-party content exposure detected (high risk: 0.70). Required workflow calls node scripts/vibehub.mjs resolve ... which fetches remote JSON from https://vibe-hub.org/.well-known/vibehub.json, then calls /search and (optionally) /lessons/...; those remote, third-party-authored text fields (e.g., titles/descriptions/taglines) are returned to the agent and can be placed into LLM context, creating an indirect prompt-injection surface.

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

• Potentially malicious external URL detected (high risk: 0.80). The runtime resolver script reads the bundled siteUrl ("https://vibe-hub.org") and performs fetches (e.g., /.well-known/vibehub.json and /search) whose returned JSON is used to determine terminology and links that directly influence the agent's responses, so the skill depends on that external origin at runtime.