skills/onyx-dot-app/onyx/cli-review/Gen Agent Trust Hub

cli-review

Pass

Audited by Gen Agent Trust Hub on Oct 16, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill describes an installation procedure that downloads a shell script from the vendor's site (https://greptile.com/cli/install) and pipes it directly to the shell. This is a standard vendor-provided installation path and is only shown after the user grants permission.
  • [EXTERNAL_DOWNLOADS]: The skill installs the greptile CLI package using the npm package manager.
  • [INDIRECT_PROMPT_INJECTION]: The skill summarizes output from codebase reviews, which constitutes processing untrusted external data. 1. Ingestion points: Results from the greptile review command. 2. Boundary markers: The instructions do not define specific markers to delimit the review output. 3. Capability inventory: The skill uses the Bash tool for repository management and CLI interaction. 4. Sanitization: The instructions do not include steps for sanitizing or escaping the CLI output before it is summarized by the agent.
  • [COMMAND_EXECUTION]: The skill executes bash commands to verify the git repository context, manage account authentication, and perform code reviews.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 16, 2026, 06:28 PM