git-advanced-workflows
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides instructions for processing Git history, which acts as an ingestion surface for potentially untrusted data.
- Ingestion points: Git commit messages, history, and branch names processed via commands like
git log,git show, andgit bisectinSKILL.md. - Boundary markers: No specific delimiters are mentioned for Git output.
- Capability inventory: The skill uses
git bisect runto execute scripts andnpm testas seen inSKILL.md. - Sanitization: No automated sanitization of commit data or branch names is described.
- [DYNAMIC_EXECUTION]: The skill documents the use of
git bisect runfor automated troubleshooting. - Evidence: Usage of
git bisect run ./test.shandgit bisect run npm testinSKILL.mdallows execution of local scripts based on repository state. - Context: This is a core feature of the Git toolchain used for its intended purpose of finding regressions.
Audit Metadata