using-git-worktrees
Pass
Audited by Gen Agent Trust Hub on Apr 12, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes various Git and system commands to inspect directory structures, verify Git ignore status, and manage worktrees. These operations are intended for local repository management and do not exhibit malicious behavior.\n- [EXTERNAL_DOWNLOADS]: The skill automates environment setup by invoking project-appropriate package managers (such as
npm,pip,cargo, andgo) which fetch dependencies from official registries. This behavior is standard for development automation and is triggered by the presence of local project manifest files.\n- [PROMPT_INJECTION]: The skill implements logic to read theCLAUDE.mdfile for worktree naming preferences. This constitutes an indirect prompt injection surface where external file content influences agent behavior.\n - Ingestion points:
CLAUDE.md(read viagrepinSKILL.md)\n - Boundary markers: Absent; the skill relies on keyword matching for specific configuration strings.\n
- Capability inventory: Subprocess execution for environment setup and Git operations, along with file-write access to
.gitignore.\n - Sanitization: Absent; the skill processes text based on pattern matching rather than validated schemas.
Audit Metadata