notes-audit

Warn

Audited by Gen Agent Trust Hub on Jul 24, 2026

Risk Level: MEDIUMCOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local TypeScript script via the bun runtime. The script is located at a specific path within the user's home directory: $HOME/.claude/plugins/marketplaces/cc-skills/plugins/notes-commander/scripts/notes.ts. This involves executing code that is not contained within the skill itself.
  • [DYNAMIC_EXECUTION]: The skill implements a 'self-evolving' pattern. It explicitly instructs the AI agent to rewrite the SKILL.md file's content (the 'smell table') to incorporate new patterns or correct false positives identified during execution. This self-modification of instructions can lead to unpredictable behavior if the agent incorrectly updates its own logic.
  • [DATA_EXPOSURE]: The skill is designed to access, export, and read the contents of the macOS Notes database. While the instructions advise sampling notes rather than reading all content, the process inherently involves the agent handling potentially sensitive personal data stored in the user's notes.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Jul 24, 2026, 11:24 AM
Security Audit — agent-trust-hub — notes-audit