Skip to content
Get packs

Guide · Agent skill

How to use the NanoClaw Customization Guide agent skill

Updated · written with AI from the project's own documentation and checked against our catalogue data

The NanoClaw Customization Guide is an agent skill named customize that helps an AI agent add capabilities to, or change the behavior of, a NanoClaw personal AI assistant project. This guide explains what a skill is, how to install this one, how to ask for changes in plain language, and what to check before letting it edit anything. It is written for developers and maintainers who already run their own NanoClaw instance.

NanoClaw Customization Guidenanocoai/nanoclaw · MIT

Quick start

  1. 1Download and unzip the skill. Download the customize skill ZIP from this page and unzip it into ~/.claude/skills/ or your project's .claude/skills/ folder, or upload it under Settings → Capabilities → Skills in the Claude apps.
  2. 2Confirm your NanoClaw project is ready. Make sure you have a NanoClaw checkout with pnpm, the ncl admin CLI, and a running service managed by launchd or systemd.
  3. 3Describe the change in plain language. Ask for the change you want, such as adding a messaging channel or changing trigger rules. Mention the platform or behavior involved.
  4. 4Answer the clarifying questions. The skill asks questions about the target agent group, isolation level and trigger rules before it edits anything. Answer them precisely.
  5. 5Review the changes. Check the files and commands the agent proposes, especially edits to core code or container configuration, before they run.
  6. 6Rebuild, restart and test. Run the build and restart steps the skill gives you, then follow its test guidance to confirm the change works.

What an agent skill is and what this one does

An agent skill is a folder containing a SKILL.md file with instructions that an AI agent loads when a request matches the skill's description. The agent then follows those instructions for that task.

This skill covers customizing NanoClaw. According to its description, it is used when you want to add channels such as Telegram, Slack or email input, change triggers, add integrations, modify the router, or make other customizations. It is an interactive skill: it asks questions to understand what you want before making changes.

  • Asks clarifying questions before editing anything
  • Points you to dedicated install skills such as /add-telegram or /manage-channels where they exist
  • Otherwise edits the right surface: a group's CLAUDE.md, the container configuration, or core code
  • Explains how to rebuild and restart the service afterward

Install the customize skill

Download the ZIP file for the skill from this page. Unzip it into one of two locations. Use ~/.claude/skills/ to make the skill available across your projects, or use a project's .claude/skills/ folder to limit it to that project.

If you use the Claude apps instead, open Settings → Capabilities → Skills and upload the skill there.

  • Personal install: unzip into ~/.claude/skills/
  • Project install: unzip into the .claude/skills/ folder of your NanoClaw project
  • App install: upload through Settings → Capabilities → Skills

Check the requirements before you start

The skill assumes an existing NanoClaw project. It is not a fresh-install helper. Before you ask for changes, confirm that you have a NanoClaw project checkout, pnpm, the ncl admin CLI, and a running service.

You also need a way to build and restart the service. The skill refers to launchd on macOS and systemd on Linux. If you add a messaging platform, you will need that platform's account or API credentials. The skill references OneCLA for credential routing.

  • A NanoClaw project checkout with pnpm and the ncl admin CLI
  • Tooling to build and restart the service: launchd on macOS or systemd on Linux
  • Channel-specific accounts or API credentials when adding a messaging platform
  • Runs scripts: yes. Uses network: yes.

How to trigger the skill with natural requests

You do not need a special command to start the skill. Describe the change in plain language and mention the kind of change you want. Requests that match the skill's description include adding a channel, changing triggers, adding an integration, or modifying the router.

The agent will then ask questions. For a new channel, expect questions such as which platform it is, whether it should reach an existing agent group or a new one, and whether it should share an isolation level with other channels.

  • Add Telegram as an input channel
  • Change the trigger rules so the assistant only responds when addressed by name
  • Add an MCP tool that gives the assistant access to my calendar
  • Change the persona of my family group's assistant
Add Telegram as an input channel

Safety notes and after-change steps

The skill can modify core code and restart services. Review its changes before they reach any production system, and do not let configuration changes go live without a check. The skill's own guidance is to tell you how to verify changes after making them.

After changes, the skill runs the build and restart sequence from your project root. On macOS it unloads and reloads the launchd agent. On Linux the restart line is written as a systemd user restart command.

  • Keep changes under version control so you can review and revert them
  • Test in a non-production instance first where possible
  • Check any network calls and credential handling the skill proposes
pnpm run build source setup/lib/install-slug.sh # macOS: launchctl unload ~/Library/LaunchAgents/$(launchd_label).plist launchctl load ~/Library/LaunchAgents/$(launchd_label).plist # Linux: # systemctl --user restart $(systemd_unit)

What people use it for

  • Adding Telegram, Slack or another messaging channel by running the matching install skill and then wiring it to an agent group
  • Changing which messages wake an agent group by updating its trigger rules
  • Wiring an MCP tool into one agent group so it can use an external service
  • Editing a group's CLAUDE.md to change the assistant's persona or standing instructions

Frequently asked questions

Is the NanoClaw customize skill free and what is its licence?

The skill is published under the MIT licence. This guide does not state any price, so check the source repository for current terms of use. Using the skill still requires your own NanoClaw project and any accounts or credentials your chosen platform needs.

How do I install the NanoClaw customize skill?

Download the ZIP from this page and unzip it into ~/.claude/skills/ for all projects, or into a project's .claude/skills/ folder for one project. In the Claude apps, you can upload it under Settings → Capabilities → Skills.

Do I need NanoClaw already installed to use this skill?

Yes. The skill assumes an existing NanoClaw project with pnpm, the ncl admin CLI and a running service. It will not help with a fresh install.

Can the skill add Telegram or Slack to NanoClaw?

It can guide you through it. The skill points you to dedicated install skills such as /add-telegram and /add-slack, which fetch the adapter, wire its registration and build. You then use /manage-channels or the ncl commands to wire the channel to an agent group.

Why won't the agent edit my code without asking questions?

The skill is designed to be interactive. Its instructions say to use clarifying questions to understand the request before making any changes, so the edit lands on the right surface.

Does the skill restart my service and is that safe?

It can. The skill can modify core code and restart services, and it lists build and restart steps for launchd on macOS and systemd on Linux. Review changes first, especially on production systems.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

Use this skill if you run your own NanoClaw instance and want guided, question-driven changes such as adding a messaging channel or adjusting trigger rules. Pick something else, or do not use it, if you have not set up NanoClaw yet or if your production system cannot accept unreviewed changes to code and services.

Download the skill