mirror of
https://github.com/pocketpaw/pocketpaw.git
synced 2026-05-21 09:14:59 +00:00
57b807c117
Comprehensive SEO optimization across 80 documentation pages: Title optimization (all pages): - Replaced generic titles like "Architecture", "Discord", "Slack" with search-intent titles like "PocketPaw Architecture: Event-Driven Message Bus", "Discord Bot Setup: Add PocketPaw to Your Server" - All titles now 50-70 characters with qualifying keywords Meta descriptions: - Expanded 7 short descriptions (under 145 chars) to 150-160 chars - Roadmap description expanded from 76 to 196 chars - Troubleshooting, Codex CLI, OpenCode, WebMCP all expanded H1 heading fixes: - Ensured single H1 per page matching the frontmatter title - All H1 headings updated to match new optimized titles Internal cross-links: - Added Related CardGroup sections to 60+ individual pages - Each links to 2-3 related pages within and across sections - Channels link to channel guides, backends link to Ollama guide, etc. Em dash cleanup: - Replaced em dashes with colons, periods, or double hyphens across multiple files in tools/, channels/, integrations/ Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
131 lines
4.2 KiB
Plaintext
131 lines
4.2 KiB
Plaintext
---
|
|
title: "Mem0 Integration: Semantic Long-Term Memory"
|
|
description: "Integrate Mem0 for semantic long-term memory with auto-learning: PocketPaw automatically extracts and stores facts from conversations, then retrieves relevant memories via vector similarity search."
|
|
section: Memory
|
|
ogType: article
|
|
keywords: ["mem0", "semantic memory", "auto-learn", "vector search", "long-term memory", "embeddings"]
|
|
tags: ["memory", "mem0", "ai"]
|
|
---
|
|
|
|
# Mem0 Integration: Semantic Long-Term Memory
|
|
|
|
[Mem0](https://mem0.ai) provides semantic memory capabilities — auto-learning from conversations and retrieving relevant memories via vector search.
|
|
|
|
## How It Works
|
|
|
|
1. **Auto-learn**: After each agent response, the conversation is sent to Mem0 in a background task
|
|
2. **Memory extraction**: Mem0's LLM extracts key facts and preferences
|
|
3. **Embedding**: Facts are embedded and stored in a vector database
|
|
4. **Retrieval**: When building context for new messages, relevant memories are retrieved via semantic search
|
|
|
|
## Setup
|
|
|
|
```bash
|
|
# Install PocketPaw
|
|
curl -fsSL https://pocketpaw.xyz/install.sh | sh
|
|
|
|
# Or add the memory extra manually
|
|
pip install pocketpaw[memory]
|
|
|
|
# Configure Mem0
|
|
export POCKETPAW_MEM0_AUTO_LEARN=true
|
|
export POCKETPAW_MEM0_LLM_PROVIDER="ollama"
|
|
export POCKETPAW_MEM0_LLM_MODEL="llama3.2"
|
|
export POCKETPAW_MEM0_EMBEDDER_PROVIDER="ollama"
|
|
export POCKETPAW_MEM0_EMBEDDER_MODEL="nomic-embed-text"
|
|
export POCKETPAW_MEM0_VECTOR_STORE="qdrant"
|
|
```
|
|
|
|
## Provider Configuration
|
|
|
|
### LLM Provider
|
|
|
|
The LLM is used by Mem0 to extract facts from conversations:
|
|
|
|
| Provider | Config | Notes |
|
|
|----------|--------|-------|
|
|
| Ollama | `ollama` / `llama3.2` | Free, local, recommended |
|
|
| Anthropic | `anthropic` / `claude-sonnet-4-5-20250929` | Needs API key |
|
|
| OpenAI | `openai` / `gpt-4o-mini` | Needs API key |
|
|
|
|
### Embedder Provider
|
|
|
|
The embedder converts text to vectors for semantic search:
|
|
|
|
| Provider | Config | Notes |
|
|
|----------|--------|-------|
|
|
| Ollama | `ollama` / `nomic-embed-text` | Free, local |
|
|
| OpenAI | `openai` / `text-embedding-3-small` | Needs API key |
|
|
|
|
### Vector Store
|
|
|
|
| Store | Description |
|
|
|-------|-------------|
|
|
| Qdrant | Default, runs embedded or as service |
|
|
| Chroma | Alternative, runs embedded |
|
|
|
|
## Embedding Dimensions
|
|
|
|
Embedding dimensions must match between the model and the vector store collection. PocketPaw includes a built-in dimensions lookup and auto-detection for Ollama models:
|
|
|
|
```python
|
|
_EMBEDDING_DIMS = {
|
|
"nomic-embed-text": 768,
|
|
"mxbai-embed-large": 1024,
|
|
"text-embedding-3-small": 1536,
|
|
"text-embedding-3-large": 3072,
|
|
}
|
|
```
|
|
|
|
For unknown Ollama models, PocketPaw queries the model to auto-detect dimensions.
|
|
|
|
## Dashboard Configuration
|
|
|
|
The web dashboard provides a Memory settings panel where you can:
|
|
|
|
- Enable/disable auto-learning
|
|
- Select LLM and embedder providers
|
|
- Choose models
|
|
- Test memory retrieval
|
|
|
|
<Callout type="warning">
|
|
After changing memory settings, you must restart the memory manager. The dashboard handles this via `get_memory_manager(force_reload=True)`. Note that `get_settings()` is LRU-cached — the cache must be cleared after saving settings.
|
|
</Callout>
|
|
|
|
## Fully Local Setup
|
|
|
|
For a completely local memory system with no external API calls:
|
|
|
|
```bash
|
|
# Install Ollama
|
|
curl -fsSL https://ollama.com/install.sh | sh
|
|
|
|
# Pull models
|
|
ollama pull llama3.2
|
|
ollama pull nomic-embed-text
|
|
|
|
# Configure
|
|
export POCKETPAW_MEM0_AUTO_LEARN=true
|
|
export POCKETPAW_MEM0_LLM_PROVIDER="ollama"
|
|
export POCKETPAW_MEM0_LLM_MODEL="llama3.2"
|
|
export POCKETPAW_MEM0_EMBEDDER_PROVIDER="ollama"
|
|
export POCKETPAW_MEM0_EMBEDDER_MODEL="nomic-embed-text"
|
|
export POCKETPAW_MEM0_VECTOR_STORE="qdrant"
|
|
```
|
|
|
|
This runs everything locally — LLM inference, embedding, and vector storage.
|
|
|
|
## Related
|
|
|
|
<CardGroup>
|
|
<Card title="File Store" icon="lucide:database" href="/memory/file-store">
|
|
PocketPaw's default JSON-based memory backend for session storage.
|
|
</Card>
|
|
<Card title="Context Building" icon="lucide:layers" href="/memory/context-building">
|
|
How semantic memories are retrieved and injected into the agent's context.
|
|
</Card>
|
|
<Card title="Memory Isolation" icon="lucide:shield" href="/memory/memory-isolation">
|
|
Per-user memory scoping when Mem0 is used in shared channels.
|
|
</Card>
|
|
</CardGroup>
|