skills.rest
LIMITLESS-LONGEVITY
LIMITLESS-LONGEVITYOFFICIAL
Explore all skills by this author →
WebsiteGitHub

add-ollama-tool

Exposes local Ollama models as callable MCP tools for containerized agents.

Updated Apr 2, 2026
Search Tags:#devops#mcp#ollama#model-management#local-models#agent-integration#container-agent
One-click install
npx skills add https://github.com/LIMITLESS-LONGEVITY/limitless --skill add-ollama-tool-limitless-longevity
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: add-ollama-tool
Source: https://github.com/LIMITLESS-LONGEVITY/limitless/tree/main/apps/nanoclaw/.claude/skills/add-ollama-tool
Command: npx skills add https://github.com/LIMITLESS-LONGEVITY/limitless --skill add-ollama-tool-limitless-longevity

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill exposes on-host Ollama models to a containerized agent so the agent can perform low-latency, private inference and optionally manage the model library without requiring remote LLM calls.

Core Features & Use Cases

  • Local inference tooling: Adds stdio-based MCP tools such as ollama_list_models and ollama_generate so the agent can discover and invoke local Ollama models.
  • Optional model lifecycle management: When enabled via OLLAMA_ADMIN_TOOLS, the agent can pull, delete, and inspect models and list running models in memory.
  • Use case: Run inference on small on-host models for private queries or offload specific tasks from a remote orchestrator to reduce latency and host bandwidth.

Quick Start

Ask the agent to use ollama_generate with a model like gemma3:1b to answer a simple question after merging the skill branch, rebuilding the container, and restarting the service.

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 a containerized agent?▼

To expose local Ollama models, you merge the skill branch, rebuild the container, and restart the service so the agent-runner registers the Ollama MCP stdio server. This allows the containerized agent to invoke local models for inference.

Can I manage the Ollama model lifecycle directly from my containerized agent?▼

Yes, you can manage the Ollama model lifecycle by enabling the OLLAMA_ADMIN_TOOLS flag. This allows the agent to pull, delete, inspect models, and list running models in memory alongside generating inference.

Does this Ollama agent integration work with a custom OLLAMA_HOST configuration?▼

Yes, this Ollama agent integration works with a custom OLLAMA_HOST configuration. It satisfies requirements for a reachable Ollama daemon, defaulting to host.docker.internal if a custom host is not specified.

Why use local Ollama models for agent inference instead of remote LLM calls?▼

Use local Ollama models for agent inference to achieve low-latency, private queries without remote LLM calls. This offloads specific tasks from a remote orchestrator, reducing latency and saving host bandwidth.

What MCP tools are added when integrating Ollama with a container agent?▼

The integration adds stdio-based MCP tools such as ollama_list_models and ollama_generate. These tools let the agent discover available local models and invoke them to answer queries.

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
theowenyoungtheowenyoung

add-ollama-tool

Expose local Ollama models to container agents via MCP tools.

Community
Advanced
VincentChong123VincentChong123

add-ollama-tool

Integrate local Ollama models as MCP tools in containerized agents.

Community
Advanced
datathingsdatathings
10

llamacpp

Reference the llama.cpp C API for model loading and inference workflows.

Official
Advanced