prowler-commit

Pass

Audited by Gen Agent Trust Hub on Oct 2, 2026

Risk Level: SAFE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill analyzes local repository data, such as git logs and diffs, to generate commit messages. While this data represents an external input surface, the risk is mitigated by the instruction to always present the proposed message to the user for explicit confirmation before execution.
  • [COMMAND_EXECUTION]: The skill utilizes standard Git commands (status, diff, log, add, commit) to perform its primary function. It includes safety rules that explicitly forbid the use of destructive flags like --force and requires user consent before making changes to the repository state.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 2, 2026, 04:39 PM