optimus:permissions

What problem does it solve?

This Skill prevents Claude Code agents from performing dangerous actions, like deleting files outside your project or pushing directly to protected branches, while still allowing them to work efficiently on tasks within your project.

Core Features & Use Cases

• Granular Control: Define exactly which tools Claude Code can use and under what conditions.
• Safety Guardrails: Automatically block destructive commands (e.g., rm -rf /, git push --force) and prevent writes outside the project directory.
• Streamlined Workflow: Eliminate routine permission prompts for safe, common operations, speeding up AI-assisted development.
• Use Case: When an AI agent needs to modify project files, this skill ensures it only edits files within the project scope and prompts for confirmation if it attempts to write to an external location. It also prevents accidental deletion of critical project files.

Quick Start

Run the optimus permissions skill to configure your project's security settings.