Skip to content
Get packs

Ollama Local Models Bridge

Lets the agent offload prompts to locally running Ollama models, and optionally manage them.

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

What it does

This skill connects an AI agent running inside a NanoClaw container to Ollama, a program that runs open AI models on your own computer. Once installed, the agent can list the models you have, send them prompts and get answers back, so cheaper or private local models can handle some work while Claude coordinates. An optional setting also lets the agent download, delete, inspect, and check the memory use of models.

Great for

  • Developers who run NanoClaw and already use Ollama on their machine
  • People who want to keep some prompts on local hardware instead of a cloud model
  • Users comparing local model output against Claude for simple tasks

Skip it if

  • Anyone without Ollama installed and running, since the tools will fail without the local daemon
  • Users who want the agent to manage disk space carelessly: enabling OLLAMA_ADMIN_TOOLS lets it delete models and pull large downloads
  • Shared or multi-user machines where other people could reach the local Ollama daemon

Try asking

  • “Use the local Ollama model gemma3:1b to summarise this text”
  • “List which Ollama models are installed on my machine”
  • “Pull the llama3.2 model from Ollama and show its details”
  • “Check which Ollama models are currently loaded in memory”
Step-by-step guide: how to use Ollama Local Models BridgeInstall, 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-ollama-tool.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

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

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip add-ollama-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

  • Only enable OLLAMA_ADMIN_TOOLS=true if you want the agent able to download or delete models; leave it unset otherwise
  • Start with a small model such as gemma3:1b to confirm the connection before using larger ones
  • Check that the Ollama daemon answers on the host before asking the agent to use it

What's inside

  • ollama-env.ts
  • ollama-mcp-stdio.ts
  • ollama-registration.test.ts
  • ollama-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.