git-worktree-create

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: LOWCOMMAND_EXECUTIONPROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [Command Execution] (SAFE): The script scripts/create_worktree.sh performs standard Git and filesystem operations.\n
  • Evidence: Uses git worktree, git fetch, mkdir, and cp for repository management.\n
  • Safety: Variables such as $BRANCH_NAME and $WORKTREE_PATH are consistently double-quoted to prevent command injection or shell expansion issues.\n- [Indirect Prompt Injection] (LOW): The skill processes external data from Git branches which could contain malicious instructions for downstream tasks.\n
  • Ingestion points: The BRANCH_NAME parameter and the content of the branch being checked out into the worktree (scripts/create_worktree.sh).\n
  • Boundary markers: Absent. The skill does not explicitly warn the agent about untrusted content in the new worktree.\n
  • Capability inventory: Performs filesystem writes (git worktree add, cp) to create the environment (scripts/create_worktree.sh).\n
  • Sanitization: Branch names are sanitized (${BRANCH_NAME//\//-}) to prevent directory traversal attacks (scripts/create_worktree.sh).\n- [External Downloads] (SAFE): The skill performs Git network operations to fetch updates.\n
  • Evidence: Uses git fetch --all to synchronize with remote repositories.\n
  • Safety: Operations are restricted to standard Git protocol and configured remotes.
Audit Metadata
Risk Level
LOW
Analyzed
Feb 17, 2026, 08:07 AM