setup-ts-deep-modules
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill identifies the project's package manager (npm, yarn, pnpm, or bun) and uses it to install the
dependency-cruiserpackage as a development dependency. - [COMMAND_EXECUTION]: The skill executes the
depcruiseCLI tool to verify the enforcement of module boundaries. This includes a verification step where the agent intentionally introduces a violation in a test file to confirm the linter correctly fails, then reverts the change. - [INDIRECT_PROMPT_INJECTION]: The skill reads existing project configuration files to adapt its behavior, creating a standard integration surface.
- Ingestion points: Reads
package.jsonto detect existing scripts and the project's package manager; scans the file system to locate the packages root. - Boundary markers: Boundary markers are not used when reading these files as they are processed as structured project data.
- Capability inventory: The skill can install packages, write configuration files (
.dependency-cruiser.cjs), updatepackage.jsonscripts, and modify agent instruction files (CLAUDE.mdorAGENTS.md). - Sanitization: The skill uses standard file system operations and structured data handling to process project information.
- [SAFE]: The skill's operations, including configuration file creation and documentation updates, are strictly aligned with its stated purpose of improving codebase architecture through automated linting.
Audit Metadata