ragclaw-plugin-ollama

RagClaw plugin for local embeddings via Ollama.

Delegates embedding to a locally-running Ollama server instead of bundling an ONNX model. Any model served by Ollama can be used to embed documents in a RagClaw knowledge base.

Requirements

• Ollama installed and running (ollama serve)
• At least one embedding model pulled (e.g. ollama pull nomic-embed-text)

Installation

npm install -g ragclaw-plugin-ollama

Configuration

Add the plugin to your RagClaw config (~/.config/ragclaw/config.yaml):

enabledPlugins:
  - ragclaw-plugin-ollama

embedder:
  plugin: ragclaw-plugin-ollama
  model: nomic-embed-text     # any model available in your Ollama server
  baseUrl: http://localhost:11434  # default; omit if using the default port

Supported Models

Dimensions are pre-wired for these popular models — no auto-detection needed:

| Model | Dimensions | |-------|-----------| | nomic-embed-text | 768 | | mxbai-embed-large | 1024 | | all-minilm | 384 | | snowflake-arctic-embed | 1024 | | bge-large | 1024 | | bge-base | 768 | | bge-small | 384 |

Any other model served by Ollama is also supported — dimensions are auto-detected on the first embedding call.

Usage

Once configured, use RagClaw as normal — the Ollama embedder is used transparently:

# Index files using the Ollama embedder
ragclaw add ./docs

# Search
ragclaw search "how does authentication work"

# Check embedder info
ragclaw status

Development

cd plugins/ragclaw-plugin-ollama
pnpm install
pnpm build
pnpm test

License

MIT