skills/gupsammy/claudest/commit/Gen Agent Trust Hub

commit

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes standard Git commands (git status, git add, git commit, git push) and project-specific linters (cargo fmt, npm run lint, ruff check). These operations are consistent with the skill's stated purpose of managing commits and validating code quality.
  • [REMOTE_CODE_EXECUTION]: The skill invokes a local Python script (scripts/validate.py) to run validation tools. This script uses subprocess.run with hardcoded command lists, which prevents command injection from external input. No remote script downloading or execution was found.
  • [DATA_EXFILTRATION]: While the skill can push code to remote repositories using git push, this action is explicitly gated by user request/intent and is a standard part of the Git workflow.
  • [SAFE]: The skill implements best practices such as conventional commit formatting, logical change grouping, and automatic exclusion of temporary or sensitive files (e.g., .log, node_modules/, build/) from being staged.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 03:40 PM