skills/laststance/skills/git/Gen Agent Trust Hub

git

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes standard local git commands including status, diff, log, commit, push, pull, branch, merge, and stash. These operations are appropriate for a version control tool and follow best practices, such as requiring confirmation for destructive actions.\n- [PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection as it processes untrusted data from the repository (via git diff) to generate commit messages.\n
  • Ingestion points: Untrusted data from the local repository is ingested via git diff, git diff --cached, and git log (SKILL.md).\n
  • Boundary markers: The skill lacks explicit boundary markers to separate the diff content from agent instructions, though it uses HEREDOC for the final command execution.\n
  • Capability inventory: The skill can perform file staging, committing, and remote push/pull operations.\n
  • Sanitization: No specific sanitization or filtering of the diff output is performed prior to context injection.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 03:29 AM
Security Audit — agent-trust-hub — git