self-customize

Automate agent customization with guided feature additions, tool installs, and code edits.

Updated Apr 26, 2026
One-click install
npx skills add https://github.com/wrightchin/nanoclaw-playground --skill self-customize-wrightchin
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: self-customize
Source: https://github.com/wrightchin/nanoclaw-playground/tree/main/container/skills/self-customize
Command: npx skills add https://github.com/wrightchin/nanoclaw-playground --skill self-customize-wrightchin

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Customizing an agent can be complex and error-prone, requiring careful decision-making about when to modify workspace, install system packages, or delegate to a builder.

Core Features & Use Cases

  • Guided decision workflow to determine how changes should be applied (workspace edits, admin installs, MCP wiring, or builder delegation).
  • Builder-agent workflow for safe, auditable code changes with explicit acceptance criteria.
  • Clear boundaries and governance (admin approvals for system changes, and safety checks to prevent unsafe operations).

Quick Start

Tell me what feature you want added, which tool or package to install, or which MCP server to wire up, and I will guide you through the appropriate steps.

Frequently Asked Questions about self-customize

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I add new features and install tools to customize my agent?

To customize your agent, a guided decision workflow determines whether changes require workspace edits, admin package installs, MCP server wiring, or builder delegation, ensuring safe feature additions. It automates these end-to-end agent customization tasks while enforcing governance.

What is the safest way to wire up an MCP server for my agent?

Wiring up an MCP server is handled through a guided workflow that separates workspace edits from container-level alterations and enforces admin approvals for system changes. This process includes safety checks to prevent unsafe operations during the server installation.

When do I need admin approval to modify agent capabilities?

Admin approval is required for agent customization whenever system-level changes are necessary, such as installing system packages or altering container configurations. The workflow clearly separates these from standard workspace edits to enforce governance and safety checks.

Can I use a builder agent to make code modifications to my agent?

Yes, a builder-agent workflow is available for safe and auditable code changes with explicit acceptance criteria. This delegates complex code modifications during agent customization, ensuring changes are reviewed and applied correctly within your workspace.

What are the limitations when editing CLAUDE-related files for agent customization?

Editing CLAUDE-related files is constrained by safety checks that prevent unsafe operations and separate workspace edits from container-level alterations. If changes require system-level access, the workflow enforces admin approvals before proceeding with the modification.

Does this agent customization workflow support installing system packages?

Yes, installing system packages is supported through an admin approval process that separates these container-level alterations from standard workspace edits. The guided workflow ensures package installations are auditable and safe before applying changes to your agent.