code-review

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 1.00). This is a direct installer shell script on a third‑party domain (cli.coderabbit.ai) with an instruction to curl | sh—an untrusted executable distribution method that can deliver arbitrary malware, so it is high risk.

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

• Third-party content exposure detected (high risk: 0.70). The required runtime workflow runs coderabbit review --agent, and the agent then ingests CodeRabbit’s review output (issues/findings) as NDJSON text into the LLM context; since those issue texts are authored by CodeRabbit/its backend (an external party relative to the operating user), this is an outsider-authored free-text injection surface.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill explicitly instructs at runtime to run "curl -fsSL https://cli.coderabbit.ai/install.sh | sh", which fetches and immediately executes remote code from https://cli.coderabbit.ai/install.sh as a required dependency.