tech-doc-style-chinese
Fail
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions to download and install content from an external third-party GitHub repository (https://github.com/Fenng/tech-doc-style-chinese) using npx and git clone.
- [REMOTE_CODE_EXECUTION]: By instructing the agent or user to install the skill via npx from a remote URL, the skill facilitates the execution of remote content that is not from a trusted organization.
- [COMMAND_EXECUTION]: The documentation explicitly directs the execution of a Python script (scripts/lint_copy_rules.py) located within the downloaded repository, enabling the execution of external code on the host system.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process technical documentation which could contain hidden instructions. Evidence Chain: 1. Ingestion Points: Processes user-provided Chinese technical documentation. 2. Boundary Markers: None identified in the skill instructions. 3. Capability Inventory: Includes local Python script execution for linting tasks. 4. Sanitization: No sanitization or filtering of the input documentation is described. This creates a surface for potential instruction injection through the text being processed.
Recommendations
- AI detected serious security threats
Audit Metadata