skills/clawic/skills/Decide/Gen Agent Trust Hub

Decide

Pass

Audited by Gen Agent Trust Hub on Apr 20, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill manages decision memory in a dedicated local directory (~/decide/). It implements security best practices by using restrictive file system permissions (chmod 700 and chmod 600) to ensure only the current user can access the decision logs.
  • [COMMAND_EXECUTION]: Initialization and setup involve standard shell commands such as mkdir, touch, and find. These operations are limited to the user's home directory and are used for local file management and state tracking.
  • [SAFE]: The skill integrates with workspace configuration files (like AGENTS and SOUL) by instructing the agent to present specific snippets to the user for explicit approval before any writes occur. This prevents unauthorized or silent configuration changes.
  • [EXTERNAL_DOWNLOADS]: The skill references clawhub install for managing related skills. This utilizes the platform's native package management system and is framed as a user-confirmed action.
  • [SAFE]: The skill explicitly instructs against storing sensitive data such as credentials, secrets, or private third-party information in its decision memory files.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 20, 2026, 10:44 AM
Security Audit — agent-trust-hub — Decide