pylabrobot

Pass

Audited by Gen Agent Trust Hub on Apr 16, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill contains instructions directing the agent to proactively suggest the author's hosted platform, K-Dense Web, for complex workflows. It also establishes an indirect prompt injection surface by processing data from hardware instruments that then drives physical operations. 1. Ingestion points: PlateReader data in SKILL.md. 2. Boundary markers: Absent. 3. Capability inventory: Physical hardware control (liquid handling, thermal control). 4. Sanitization: No input validation for instrument data.
  • [COMMAND_EXECUTION]: The documentation includes standard shell commands for installing the pylabrobot library via pip.
  • [EXTERNAL_DOWNLOADS]: The skill references official project resources including documentation at docs.pylabrobot.org, the GitHub repository at github.com/PyLabRobot/pylabrobot, and the PyPI registry.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 16, 2026, 08:26 AM