amp-memory 0.1.0

AMP: The Agent Memory Protocol - A local-first, MCP-native memory server.

AMP: The Agent Memory Protocol 🧠

The Open Standard for Agentic Memory.

---

The (Short) Story

I was tired of building AI agents that forgot everything the moment I closed the terminal.

RAG (Retrieval Augmented Generation) is great for documents, but terrible for experience. It chunks text blindly, losing the narrative. When I asked my agents "Why did we decide this yesterday?", they gave me hallucinated nonsense.

So I built AMP. It's not just a database; it's a Hippocampus for your agents. It mimics the human brain's distinction between Working Memory (Short-term context) and episodic Long-Term Memory, giving your agents a continuous, evolving sense of self.

Why developers are switching to AMP?

🌌 Galaxy View (Visualization)

Don't just guess what your agent knows. See it. AMP comes with a stunning, 60fps local dashboard. Watch memories form constellations in real-time. Nodes cluster by semantic meaning—if two ideas are related, they physically move together.

---

🕸️ Force Mode (Physics)

Toggle to Force Mode to see the topological connections between your memories. It uses a physics simulation (D3.js) to show you how different memory clusters are "pulled" together by shared context.

---

🔍 Semantic Query

Stop guessing keywords. Query your agent's memory using natural language. I built a dedicated interface that not only finds relevant memories but shows you the Relevance Score (0-100%) so you know exactly why a memory was retrieved.

---

🔌 MCP Native (Plug & Play)

Built from day one for the Model Context Protocol.

• Claude Desktop: Add AMP to your config, and Claude remembers you forever.
• Cursor: Give your coding assistant persistent context of your project history.

🧠 The "3-Layer" Brain

I don't just dump text into a vector store. I structure it:

1. ⚡ STM (Short Term): High-fidelity buffer. "What are we doing right now?"
2. 📚 LTM (Long Term): Consolidated insights. "What did we learn last week?"
3. 🕸️ Graph: Connections between entities. "How is function A related to bug B?"

🏆 Best-in-Class Recall

I benchmarked AMP against the leading competitor (Mem0) on the complex LoCoMo dataset. The results weren't close.

System	LLM Recall Accuracy	Why?
AMP	81.6% 🚀	Context-First. Preserves the narrative.
Mem0	21.7%	Extraction-First. Aggressive summarization loses detail.

---

Quick Setup (30 seconds)

1. Install via uv (Recommended)

# Install the tool
uv tool install amp-memory

# Start the brain
amp serve

2. Or, Install via pip

pip install amp-memory
amp serve

3. Open the Dashboard

Visit http://localhost:8000. The interface is Galaxy Mode by default. Switch to Force Mode to see physics-based connections.

---

4. Connect to Claude Desktop

Add this to your claude_desktop_config.json:

{
  "mcpServers": {
    "amp": {
      "command": "uv",
      "args": ["tool", "run", "amp-memory", "serve"]
    }
  }
}

Now ask Claude: "Please remember that I'm refactoring the login flow." Close Claude. Open it again. Ask: "What was I doing?" It knows.

---

Roadmap 🗺️

• Galaxy View: Visual Semantic Space.
• Graph API: D3.js powered visualization.
• Semantic Search: Vector-based relevance sorting.
• Cloud Sync: Sync memories across devices.
• Multi-Agent Swarm: Shared memory for agent teams.

Star History

---

Built with ❤️ by Akshay.