Skip to content
Get packs

Atomic Chat Local Models

Lets the NanoClaw container agent call local models served by the Atomic Chat desktop app

Source Download skill
Repository
nanocoai/nanoclaw
Stars
31,072
Licence
MIT
Files
6 files

What it does

This skill adds a small tool server to the NanoClaw agent so that Claude can hand some tasks to AI models running on your own Mac through the Atomic Chat desktop app. Claude stays in charge, but it can list the models you have installed and send them prompts, getting answers back from the local machine instead of a cloud service. Installing it means copying files into the NanoClaw code and editing several of its source files.

Great for

  • Developers running NanoClaw on macOS who already use Atomic Chat for local models
  • People who want cheaper or private offloading of simple prompts to a local model

Skip it if

  • Users without Atomic Chat installed, since the tools will fail without its local API on port 1337 (macOS only per the instructions)
  • Anyone unwilling to let a skill modify core NanoClaw source files and Docker-related logging code, since it edits several integration points and a shared logger block

Try asking

  • “Add the Atomic Chat integration to my NanoClaw agent”
  • “List the models currently available in Atomic Chat and tell me which is best for code review”
  • “Send this summary prompt to my local Qwen coder model in Atomic Chat and return the answer”
Step-by-step guide: how to use Atomic Chat Local ModelsInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 6 original files, the MIT licence and an attribution note.

    add-atomic-chat-tool.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip add-atomic-chat-tool.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip add-atomic-chat-tool.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Run the verification commands in Phase 2 and confirm they pass before using the tools.
  • Keep Atomic Chat running with a model loaded; the skill only talks to it over the local API, so nothing is sent to a cloud service by this skill itself.
  • Review the shared [ATOMIC] logger edit in docker-driver.ts, since other local-model skills also edit that block.

What's inside

  • atomic-chat-env.ts
  • atomic-chat-mcp-stdio.ts
  • atomic-chat-registration.test.ts
  • atomic-chat-wiring.test.ts
  • REMOVE.md
  • SKILL.md

Redistributed unmodified from nanocoai/nanoclaw under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.