using-git-worktrees
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill automatically acts upon files in a repository which could contain malicious instructions or code.
- Ingestion points: The skill reads project manifests like
package.json,Cargo.toml,requirements.txt,pyproject.toml,go.mod, and metadata fromCLAUDE.mdto drive its automated actions. - Boundary markers: No delimiters or safety instructions are used to prevent the agent from obeying instructions hidden within these project files.
- Capability inventory: The skill uses shell commands (in
SKILL.md) to run package managers (npm,pip,poetry,cargo,go) and test suites, which effectively execute third-party code defined in the repository. - Sanitization: There is no evidence of validation or sanitization of the manifest file contents before the agent executes the corresponding build or installation commands.
- [EXTERNAL_DOWNLOADS]: Automates the process of fetching dependencies from established public registries such as NPM, PyPI, and Crates.io upon detecting relevant manifest files.
- [COMMAND_EXECUTION]: Utilizes shell commands for Git workspace management, directory navigation, and project setup tasks.
- [DYNAMIC_EXECUTION]: Triggers the runtime execution of build scripts and tests which are dynamically determined by the project's configuration environment.
Audit Metadata