ClawKit Logo
ClawKitReliability Toolkit
Back to Registry
Official Verified developer tools Safety 5/5

smart-memory

Context-aware memory for AI agents with dual retrieval modes — fast vector search or curated Focus Agent synthesis. SQLite backend, zero configuration, local embeddings.

Why use this skill?

Enhance OpenClaw with Smart Memory. A local, privacy-focused tool offering hybrid vector and keyword search with advanced Focus Agent synthesis for complex queries.

skill-install — Terminal

Install via CLI (Recommended)

clawhub install openclaw/skills/skills/getmolty/smart-memorycn
Or

What This Skill Does

Smart Memory v2.1 acts as a powerful, drop-in replacement for the standard OpenClaw memory system. It transforms how your agent interacts with historical data by leveraging a sophisticated hybrid search engine. Built upon a robust SQLite backend, it utilizes FTS5 for keyword-based BM25 ranking and local Transformers.js embeddings for semantic understanding. The key innovation is the 'Focus Agent' feature—a multi-pass retrieval system that doesn't just return raw data, but retrieves, ranks, and synthesizes information into a coherent, confidence-scored narrative for the user. Because it operates entirely locally, it maintains maximum privacy while removing the need for external database dependencies or API keys for vector generation.

Installation

To get started with Smart Memory, ensure you have your OpenClaw environment initialized. Open your terminal and run the following command to install the package directly from the hub:

clawhub install openclaw/skills/skills/getmolty/smart-memory

Once the installation finishes, initialize your local database by running the synchronization command:

node smart-memory/smart_memory.js --sync

This setup requires zero configuration, allowing the skill to begin indexing your local files and context immediately.

Use Cases

Smart Memory excels in scenarios where information density is high and accuracy matters. Use 'Fast Mode' for retrieving quick syntax references, code snippets, or routine facts. Switch to 'Focus Mode' when you are tackling high-level architectural decisions, planning projects, or need to synthesize information spread across disparate documents. It is particularly effective for teams maintaining long-term project history, as the system provides a persistent source of truth that remains localized and searchable within the agent's workspace.

Example Prompts

  1. "What was the agreed-upon design pattern for the authentication module discussed in last month's project meetings?"
  2. "Search through my technical notes to find the CLI command for the remote git sync we used during the Alpha phase."
  3. "Focus Mode: Synthesize the pros and cons of the database migration strategy we discussed last week for the main application core."

Tips & Limitations

For optimal performance, ensure your source files are consistently synced using the --sync flag after significant updates. While 'Focus Mode' provides superior, context-aware answers, it is more computationally intensive than 'Fast Mode' because it performs multi-pass ranking and narrative synthesis. Reserve 'Focus Mode' for complex queries requiring deep synthesis. Note that because this tool uses local Transformers.js models, your CPU usage may spike briefly during the embedding process, which is expected behavior for local, private AI processing.

Metadata

Author@getmolty
Stars1100
Views1
Updated2026-02-17
View Author Profile
AI Skill Finder

Not sure this is the right skill?

Describe what you want to build — we'll match you to the best skill from 16,000+ options.

Find the right skill
Add to Configuration

Paste this into your clawhub.json to enable this plugin.

{
  "plugins": {
    "official-getmolty-smart-memorycn": {
      "enabled": true,
      "auto_update": true
    }
  }
}

Tags(AI)

#memory#sqlite#vector-search#productivity#local-ai
Safety Score: 5/5

Flags: file-read, file-write