lyxy-runner-python
Pass
Audited by Gen Agent Trust Hub on Nov 7, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions for installing the 'uv' package manager via its official installation script at 'https://astral.sh/uv/install.ps1'. This is a reference to a well-known tool from its official vendor.
- [REMOTE_CODE_EXECUTION]: The skill's primary function is to execute Python scripts using 'uv run'. It establishes a protocol for declaring dependencies using PEP 723 inline metadata or command-line flags to ensure scripts run in isolated environments.
- [COMMAND_EXECUTION]: The skill utilizes shell commands to invoke 'uv', manage Python environments, and run temporary scripts. It includes safeguards such as using the system temporary directory and avoiding global package installation.
- [INDIRECT_PROMPT_INJECTION]: As a code runner, the skill has an inherent attack surface for processing untrusted content.
- Ingestion points: Reads and executes 'external scripts' (e.g., 'external_script.py') mentioned in 'SKILL.md'.
- Boundary markers: Uses 'uv' isolation but does not provide specific prompt-level delimiters for the code content itself.
- Capability inventory: Full Python execution capabilities (subprocess calls, file system access, networking) are available within the 'uv' environment.
- Sanitization: The skill relies on 'uv' for environment-level isolation and standard agent safeguards during script creation.
Audit Metadata