What is Chromadb Memory?

A self-hosted long term memory solution that uses ChromaDB and Ollama embeddings to auto recall relevant memories into each agent turn.

Do I need cloud services?

No. It runs locally with no cloud dependencies, keeping all memories on your hardware.

How does auto recall work?

Before each turn, the plugin embeds the user message with Ollama, queries ChromaDB for nearest memories, and injects top results above minScore into the context.

Chromadb Memory Pub

Verified

@msensintaffar

npx machina-cli add skill @msensintaffar/chromadb-memory --openclaw

Files (1)

SKILL.md

3.6 KB

ChromaDB Memory

Long-term semantic memory backed by ChromaDB and local Ollama embeddings. Zero cloud dependencies.

What It Does

Auto-recall: Before every agent turn, queries ChromaDB with the user's message and injects relevant context automatically
chromadb_search tool: Manual semantic search over your ChromaDB collection
100% local: Ollama (nomic-embed-text) for embeddings, ChromaDB for vector storage

Prerequisites

ChromaDB running (Docker recommended):

docker run -d --name chromadb -p 8100:8000 chromadb/chroma:latest

Ollama with an embedding model:
```
ollama pull nomic-embed-text
```
Indexed documents in ChromaDB. Use any ChromaDB-compatible indexer to populate your collection.

Install

# 1. Copy the plugin extension
mkdir -p ~/.openclaw/extensions/chromadb-memory
cp {baseDir}/scripts/index.ts ~/.openclaw/extensions/chromadb-memory/
cp {baseDir}/scripts/openclaw.plugin.json ~/.openclaw/extensions/chromadb-memory/

# 2. Add to your OpenClaw config (~/.openclaw/openclaw.json):

{
  "plugins": {
    "entries": {
      "chromadb-memory": {
        "enabled": true,
        "config": {
          "chromaUrl": "http://localhost:8100",
          "collectionName": "longterm_memory",
          "ollamaUrl": "http://localhost:11434",
          "embeddingModel": "nomic-embed-text",
          "autoRecall": true,
          "autoRecallResults": 3,
          "minScore": 0.5
        }
      }
    }
  }
}

# 4. Restart the gateway
openclaw gateway restart

Config Options

Option	Default	Description
`chromaUrl`	`http://localhost:8100`	ChromaDB server URL
`collectionName`	`longterm_memory`	Collection name (auto-resolves UUID, survives reindexing)
`collectionId`	—	Collection UUID (optional fallback)
`ollamaUrl`	`http://localhost:11434`	Ollama API URL
`embeddingModel`	`nomic-embed-text`	Ollama embedding model
`autoRecall`	`true`	Auto-inject relevant memories each turn
`autoRecallResults`	`3`	Max auto-recall results per turn
`minScore`	`0.5`	Minimum similarity score (0-1)

How It Works

You send a message
Plugin embeds your message via Ollama (nomic-embed-text, 768 dimensions)
Queries ChromaDB for nearest neighbors
Results above minScore are injected into the agent's context as <chromadb-memories>
Agent responds with relevant long-term context available

Token Cost

Auto-recall adds ~275 tokens per turn worst case (3 results × ~300 chars + wrapper). Against a 200K+ context window, this is negligible.

Tuning

Too noisy? Raise minScore to 0.6 or 0.7
Missing context? Lower minScore to 0.4, increase autoRecallResults to 5
Want manual only? Set autoRecall: false, use chromadb_search tool

Architecture

User Message → Ollama (embed) → ChromaDB (query) → Context Injection
                                                  ↓
                                          Agent Response

No OpenAI. No cloud. Your memories stay on your hardware.

Source

git clone https://clawhub.ai/msensintaffar/chromadb-memoryView on GitHub

Overview

Chromadb Memory provides long-term semantic memory using ChromaDB and local Ollama embeddings. It injects relevant recalled context into each turn, keeping sensitive data on premises with no cloud dependencies.

How This Skill Works

On each message, the plugin embeds the text with Ollama using the nomic-embed-text model. It then queries ChromaDB for nearest memories and injects the top results above a minScore into the agent context before generating a reply.

When to Use It

You want persistent memory across user chats without cloud services
You require data privacy and local hosting for sensitive tasks
You want automatic context injection before every agent turn
You want to perform manual semantic search via chromadb_search when needed
You are operating in environments with no external API access

Quick Start

Step 1: Run ChromaDB locally (docker run ) and boot Ollama with an embedding model
Step 2: Install the chromadb-memory extension and add it to your OpenClaw config with chromaUrl, collectionName, ollamaUrl, embeddingModel, autoRecall, and minScore
Step 3: Restart the gateway and start using the agent with automatic recall enabled

Best Practices

Ensure ChromaDB and Ollama are running locally before enabling the plugin
Tune minScore and autoRecallResults to balance relevance with noise
Use the default collectionName longterm_memory to survive reindexing
Experiment with chromadb_search for controlled memory retrieval
Monitor token impact per turn and adjust settings accordingly

Example Use Cases

Support agents recall user preferences and prior tickets to resolve issues faster
Researchers remember prior sources and summaries to build on previous work
Onboarding assistants remember company policies and product docs across sessions
Privacy focused assistants store and reference sensitive data entirely on premises
Knowledge workers link reminders and notes to current prompts for context

Frequently Asked Questions

Add this skill to your agents