convert-cc-defs

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly fetches and ingests files from public GitHub repos (e.g., EveryInc/compound-engineering-plugin and obra/superpowers) using gh API calls in "Phase 1: Identify and Fetch" and then runs mechanical conversion plus an LLM-driven "Intelligent rewrite" on that upstream content, meaning untrusted third‑party content is read and can materially influence tool behavior and decisions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly runs gh API calls to fetch upstream agent/skill files from the CEP GitHub repo (e.g., https://github.com/EveryInc/compound-engineering-plugin via the gh api repos/EveryInc/compound-engineering-plugin/contents/… endpoints) at runtime; those fetched CC-format files are then converted into agent/skill definitions and therefore directly control prompts/instructions—constituting a high-risk remote instruction dependency.