uv
Fail
Audited by Gen Agent Trust Hub on Feb 27, 2026
Risk Level: HIGHEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: Provides instructions for installing the
uvtool using official shell and PowerShell scripts fromastral.sh. These are standard installation patterns for this well-known utility. - [EXTERNAL_DOWNLOADS]: Fetches installation scripts and references documentation from
astral.sh, the official provider for theuvtool. - [COMMAND_EXECUTION]: Uses
uv run,uv sync, anduv addto manage environments and execute code. These commands operate within the local project environment. - [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it processes data from external project files.
- Ingestion points: The skill reads and acts upon
uv.lock,pyproject.toml, andrequirements.txtfiles (referenced in SKILL.md). - Boundary markers: No specific delimiters or instructions to ignore embedded commands within these files are provided.
- Capability inventory: The skill can execute arbitrary commands via
uv runand install new dependencies viauv addbased on project context. - Sanitization: There is no evidence of sanitization or validation of the content within the configuration files before they are processed by the agent.
Recommendations
- HIGH: Downloads and executes remote code from: https://astral.sh/uv/install.sh - DO NOT USE without thorough review
Audit Metadata