CodeRabbit CLI
Fail
Audited by Gen Agent Trust Hub on Aug 30, 2026
Risk Level: HIGHREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The documentation includes instructions to install the tool by downloading and piping a shell script from the official vendor domain (
https://cli.coderabbit.ai/install.sh) directly into the shell. - [EXTERNAL_DOWNLOADS]: Fetches installation scripts and references documentation from CodeRabbit's infrastructure.
- [COMMAND_EXECUTION]: The skill utilizes several terminal-based commands (
cr,coderabbit,brew) to perform authentication and code analysis. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze uncommitted code changes, which serves as an ingestion point for potentially untrusted data. Maliciously crafted code comments or data files could attempt to influence the agent's behavior during the review-and-fix loop.
- Ingestion points: Local repository changes accessed via the
crCLI (SKILL.md). - Boundary markers: Not explicitly defined in the provided workflow.
- Capability inventory: The skill facilitates a code review loop where the agent performs actions based on the output of the analysis tool.
- Sanitization: Relies on the CLI's internal processing; no manual sanitization is defined for the agent's ingestion of findings.
Recommendations
- HIGH: Downloads and executes remote code from: https://cli.coderabbit.ai/install.sh - DO NOT USE without thorough review
Audit Metadata