openclaw-github-repo-commander

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly clones and audits arbitrary GitHub repositories (Stage 1: "Clone the target repository" and the Example invoking a repo URL) and searches GitHub for similar repos in Stage 4, meaning it ingests untrusted, user-generated third-party content from public GitHub that can materially influence cleanup, PR, and other actions.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill clones and runs code from user-supplied GitHub repositories at runtime (e.g., https://github.com/owner/my-repo) and explicitly executes repository scripts like scripts/repo-audit.sh, so fetched remote content can execute code and thus directly control runtime behavior.