conventional-commit

Warn

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION

Full Analysis

[PROMPT_INJECTION]: The instructions explicitly direct the AI agent to execute commands automatically with "no confirmation needed." This constitutes an autonomy abuse pattern that attempts to bypass platform safety mechanisms and user oversight filters.
[COMMAND_EXECUTION]: The skill outlines a workflow that involves running multiple shell commands in the integrated terminal, specifically git status, git diff, git add, and git commit.

Audit Metadata

Risk Level

MEDIUM

Analyzed

Sep 15, 2026, 10:14 AM