github-kb

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly uses the GitHub CLI and git clone to search and pull public GitHub repos and instructs the agent to "Read README or key files" and the GITHUB_KB.md catalog (see "Searching Repos", "Cloning a New Repo to KB", and "Local Knowledge Base Workflow" in SKILL.md), which ingests untrusted, user-generated third‑party content that could contain instructions influencing agent actions.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill performs runtime git clones (git clone https://github.com//.git) and then reads README/key files from the fetched repository to generate descriptions, so externally fetched repository content can directly influence the agent's prompts/outputs.