skills.rest
wrightchin
wrightchinCOMMUNITY
Explore all skills by this author →
GitHub

add-ollama-tool

Expose local Ollama models as deterministic tools via an MCP server.

Updated Apr 26, 2026
Search Tags:#tooling#mcp#container#ollama#model-management#local-models
One-click install
npx skills add https://github.com/wrightchin/nanoclaw-playground --skill add-ollama-tool-wrightchin
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: add-ollama-tool
Source: https://github.com/wrightchin/nanoclaw-playground/tree/main/.claude/skills/add-ollama-tool
Command: npx skills add https://github.com/wrightchin/nanoclaw-playground --skill add-ollama-tool-wrightchin

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Expose local Ollama models as deterministic tools for container agents.

Core Features & Use Cases

  • Provides an stdio-based MCP server to access local Ollama models as tools.
  • Includes management options (pull, delete, show, list running) for admin users (opt-in).
  • Enables per-group agent-runner integration with automatic config and restart workflows.

Quick Start

Follow the integration steps to enable the Ollama MCP server tools in the agent environment.

Frequently Asked Questions about add-ollama-tool

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

FAQPage Schema
How do I expose local Ollama models as tools for container agents?▼

You can expose local Ollama models as deterministic tools for container agents by deploying an stdio-based MCP server. This integration enables per-group agent-runner configurations with automatic restart workflows to ensure a ready-to-use environment.

Can I manage local Ollama models like pulling or deleting through an MCP server?▼

Yes, the MCP server includes optional admin tools to manage local Ollama models. Administrators can opt-in to execute management commands such as pulling, deleting, showing, and listing running models inside the host.

How does an MCP server integrate with Ollama for agent model management?▼

The MCP server integrates with Ollama using a stdio transport layer implemented via TypeScript modules. It configures per-group agent-runners and performs restart and verify steps to ensure the local models are ready to use.

Do I need to manually restart container agents after configuring Ollama model tools?▼

No, manual restarts are not needed because the system features automatic config and restart workflows. It implements restart and verify steps to automatically ensure the agent environment is ready after applying per-group configurations.

What are the limitations of using local Ollama models as deterministic tools?▼

The primary limitation is that local Ollama models function as deterministic tools constrained by the host environment's resources. Management capabilities like pulling or deleting models are restricted to opt-in admin users only.

Related Skills

Explore More Software Engineering Skills→
jeffreykthomasjeffreykthomas

add-ollama-tool

Expose on-host Ollama models to NanoClaw agents via an MCP server.

Community
Intermediate
VincentChong123VincentChong123

add-ollama-tool

Integrate local Ollama models as MCP tools in containerized agents.

Community
Advanced
rawvegrawveg
40

ollama

Run and interact with local AI models via the Ollama API.

Community
Advanced
Mesh-LLMMesh-LLM
3.1k

connect-agents

Connect agent tools and OpenAI clients to a mesh-llm node.

Official
Intermediate