skills/iliaal/ai-skills/git-worktree/Gen Agent Trust Hub

git-worktree

Pass

Audited by Gen Agent Trust Hub on Sep 21, 2026

Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONPERSISTENCE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses a bash script to execute Git commands (git worktree, git fetch, git rev-parse) for creating and managing parallel development environments.
  • [DATA_EXPOSURE]: The worktree-manager.sh script copies sensitive .env files from the main repository to new worktree paths to ensure configuration consistency. This operation is limited to the local filesystem within the project root.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input in the form of branch names and repository manifest files. It mitigates injection risks by validating branch names with git check-ref-format and verifying worktree ownership through unique session identifiers.
  • [DYNAMIC_EXECUTION]: Instructions guide the agent to automatically run package installation commands (e.g., npm install, pip install) when standard manifest files are detected. This facilitates automated environment provisioning.
  • [PERSISTENCE]: The skill includes technical documentation on implementing Git hooks to automate tasks. While hooks are a persistence mechanism, the provided guidance focuses on safe implementation within a worktree-aware environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 21, 2026, 03:31 PM
Security Audit — agent-trust-hub — git-worktree