Skills
skills/stevengonsalvez/agents-in-a-box/global-learnings/Gen Agent Trust Hub

global-learnings

Pass

Audited by Gen Agent Trust Hub on May 5, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes system commands to manage dependencies and search indices.
  • scripts/setup-qmd.sh runs npm install -g @tobilu/qmd to install the QMD tool and brew install sqlite on macOS.
  • scripts/learnings_cli.py uses subprocess.run to call git for knowledge base versioning and qmd for embedding updates.
  • [EXTERNAL_DOWNLOADS]: The skill fetches resources from public software registries and model hubs.
  • scripts/learnings installs Python dependencies via uv pip install and downloads the all-mpnet-base-v2 embedding model from Hugging Face.
  • scripts/setup-qmd.sh installs the qmd package from the NPM registry.
  • [DATA_EXFILTRATION]: The skill modifies user shell configuration files (.zshrc, .bashrc, .bash_profile) to persist the LEARNINGS_HOME environment variable. This is a transparent configuration step required for the tool's intended system-wide functionality.
Audit Metadata
Risk Level
SAFE
Analyzed
May 5, 2026, 09:05 AM