vibe-goal

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s runtime flow ingests ticket/spec text only by fetching the full body from an issue tracker via /vibe-plan and then passes that content into per-ticket /vibe-implement, so outsider-authored free text in the tracker can reach the LLM without prior item selection by the agent.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly fetches a ticket's full body from the tracker at runtime (SKILL.md line 89), and the eval input includes the external issue URL https://github.com/acme/app/issues/120 which would be fetched and injected into sub-agent prompts, so this URL is a runtime external dependency that can directly control agent prompts.