skill-creator
Pass
Audited by Gen Agent Trust Hub on Sep 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the subprocess module to call the claude CLI for running evaluations and optimizing skill descriptions. It also uses system utilities like lsof and kill to manage local network ports for its viewer component.
- [DYNAMIC_EXECUTION]: The eval-viewer/generate_review.py script starts a local HTTP server on the loopback interface (127.0.0.1) to provide a graphical interface for reviewing evaluation outputs. It uses the webbrowser module to open the local server.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided test prompts and skill instructions during the evaluation process. 1. Ingestion points: evals/evals.json and SKILL.md. 2. Boundary markers: Uses structured JSON and instructions to separate test data from operational logic. 3. Capability inventory: Spawns subagents, performs file operations, and executes CLI commands. 4. Sanitization: Employs JSON serialization and HTML escaping for user-facing reports.
- [EXTERNAL_DOWNLOADS]: The viewer loads the SheetJS library from a well-known CDN (cdn.sheetjs.com) and fonts from Google's trusted repositories.
Audit Metadata