claude-switch-models-setup

Pass

Audited by Gen Agent Trust Hub on Jul 26, 2026

Risk Level: SAFECOMMAND_EXECUTIONCREDENTIALS_UNSAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill installs a background service on macOS using a LaunchAgent (ai.daymade.claude-skill-source-sync). This service monitors local directory-source manifests and automatically repairs symlinks between source repositories and the Claude Code/Codex runtime environments.
  • [CREDENTIALS_UNSAFE]: The system manages LLM provider API keys by storing them in plain-text JSON files within the ~/.claude/settings/ directory. The skill documentation explicitly notes this is consistent with the native security model of the Claude Code CLI.
  • [COMMAND_EXECUTION]: The setup process involves modifying the user's shell configuration (e.g., ~/.zshrc or ~/.bashrc) to add aliases and source the profile management environment.
  • [COMMAND_EXECUTION]: Python scripts (sync-profile-settings.py, claude-plugins-sync.py, sync-local-skill-sources.py) are used to perform administrative tasks such as merging configuration files, creating directory structures, and managing cross-process locks during synchronization.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 26, 2026, 05:06 AM
Security Audit — agent-trust-hub — claude-switch-models-setup