skills/wu-yc/labclaw/cosmic-database/Gen Agent Trust Hub

cosmic-database

Pass

Audited by Gen Agent Trust Hub on Mar 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill documentation describes standard programmatic access to a well-known scientific database (COSMIC). All external references point to the official COSMIC registration and help pages (cancer.sanger.ac.uk).
  • [DATA_EXPOSURE]: The documentation includes placeholders for credentials (email="your_email@institution.edu", password="your_password"), which is standard practice for teaching users how to authenticate and does not constitute a security risk.
  • [COMMAND_EXECUTION]: The skill provides standard installation commands for well-known Python libraries (requests, pandas, pysam) from official package registries.
  • [PROMPT_INJECTION]: While the skill contains a section suggesting the use of 'K-Dense Web' for complex workflows, this is a legitimate promotional instruction from the skill's author and does not attempt to bypass LLM safety guidelines or override core agent behavior.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 15, 2026, 04:31 AM