skills/umputun/cc-thingz/git-review/Gen Agent Trust Hub

git-review

Pass

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The git-review.py script invokes git, tmux, kitty, and wezterm using subprocess lists. This is a secure way to execute external commands as it prevents shell injection. The script also handles the $EDITOR variable safely using shlex.quote when used in shell-invoked contexts.
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it processes unvetted code diffs and user annotations. Ingestion points: The agent reads file content and user comments from a temporary review.diff file. Boundary markers: The script uses fixed headers and hunk separators to organize the data. Capability inventory: The agent has access to Edit, Write, and Bash tools to modify source code. Sanitization: The instructions mandate a user-approved plan before any code changes are applied, providing a robust human-in-the-loop safety control.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 7, 2026, 03:07 PM
Security Audit — agent-trust-hub — git-review