ruff
Pass
Audited by Gen Agent Trust Hub on Apr 27, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides legitimate documentation for Ruff, a standard Python developer tool.
- [EXTERNAL_DOWNLOADS]: The skill utilizes 'uvx' to download the Ruff binary from a well-known service, which is a common and expected practice.
- [REMOTE_CODE_EXECUTION]: Execution of the Ruff binary is standard functionality for linting and formatting tasks.
- [COMMAND_EXECUTION]: The instructions involve executing shell commands on local project files to perform analysis and fixes.
- [DATA_EXFILTRATION]: The 'ruff analyze graph' command analyzes code structure; however, no instructions for transmitting this data to external endpoints are present.
- [PROMPT_INJECTION]: The skill demonstrates a surface for indirect prompt injection by processing untrusted local data. 1. Ingestion points: Python files and configuration files (SKILL.md, references/configuration.md). 2. Boundary markers: Absent; instructions do not specify ignoring directives or instructions embedded in comments. 3. Capability inventory: Shell command execution of 'ruff' and 'uv' (SKILL.md). 4. Sanitization: Absent; tool output is presented directly to the agent.
Audit Metadata