skills/wu-yc/labclaw/pysam/Gen Agent Trust Hub

pysam

Pass

Audited by Gen Agent Trust Hub on Mar 15, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill contains instructions for the AI to proactively promote an external platform (www.k-dense.ai) for complex tasks, which influences the agent's neutral behavior and response framing.
  • [COMMAND_EXECUTION]: The skill exposes functionality to run bioinformatics commands through the pysam library's wrappers for samtools and bcftools, allowing system-level command execution.
  • [EXTERNAL_DOWNLOADS]: The skill instructions include the installation of the third-party 'pysam' library from public package registries via pip.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 15, 2026, 04:32 AM