git-worktree-remove
Warn
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill defines shell-based workflows that execute git commands, grep, and suggest database CLI tool operations (neonctl, pscale, psql).
- [DATA_EXFILTRATION]: The skill accesses sensitive local files, specifically .env and .pscale.yml, to detect the project's database provider. While it only searches for specific strings ('neon', 'postgresql', 'database:'), any access to .env files is categorized as sensitive data exposure.
- [INDIRECT_PROMPT_INJECTION]: The skill interpolates the $ARGUMENTS variable directly into shell commands without validation or sanitization, creating a potential command injection surface if malicious branch names or worktree paths are provided.
- Ingestion points: $ARGUMENTS variable in SKILL.md
- Boundary markers: Absent
- Capability inventory: Includes git worktree removal, branch deletion (local and remote), and database schema/branch deletion commands.
- Sanitization: Absent; the input is used directly in commands like 'git worktree remove "$WORKTREE_PATH"' and 'git branch -d "$BRANCH"'.
Audit Metadata