skills/zrong/skills/git-commit/Gen Agent Trust Hub

git-commit

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local system commands including git for repository management and python3 to run a bundled versioning script (scripts/calver.py). These operations are core to the skill's functionality.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from the local Git repository, such as output from git status, git diff, and git tag -l. This untrusted external data is used to generate commit messages and determine version numbers.
  • Ingestion points: The skill reads git status, git diff, and git tag -l (referenced in SKILL.md and scripts/calver.py).
  • Boundary markers: The skill defines a strict Commit Message template, but does not specify delimiters for wrapping raw diff content during the processing phase.
  • Capability inventory: The skill possesses the ability to modify the local file system via git commit, access the network via git push, and manipulate repository history via git rebase.
  • Sanitization: The scripts/calver.py script uses regular expressions to validate existing tags, but the instructions do not explicitly mandate sanitization for user-provided or repository-sourced strings before inclusion in commit messages.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 07:51 AM
Security Audit — agent-trust-hub — git-commit