code-research

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly clones and reads arbitrary GitHub repositories (see "Clone if missing" with gh repo clone {org}/{repo} and the agent prompt in SKILL.md instructing the code-researcher to use Read/Grep/Glob on ~/code/{org}/{repo}), meaning it ingests untrusted, user-generated third‑party content that can influence agent behavior.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly clones GitHub repos at runtime (gh repo clone https://github.com/{org}/{repo}) and spawns an agent that will read and act on the repository contents, so fetched remote repo files can directly influence the agent's prompts and behavior.