A local-first search utility designed to index personal data and provide rapid retrieval via cached memory queries.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install mem
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install mem using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Memory Search?
Memory Search is a powerful local-first indexing and retrieval tool designed for developers and AI agents who need immediate access to their stored information. Part of the Openclaw Skills ecosystem, it allows for the creation of a persistent cached index of your documents or interaction history, enabling lightning-fast lookups without relying on external cloud processing for every query.
This skill is particularly optimized for chat-based environments where speed and context are critical. By maintaining a local index, it ensures that your AI coding agent can pull relevant headers and file paths instantly, providing the necessary context to complete complex tasks or answer specific user inquiries through interfaces like Telegram.
Memory Search Use Cases
- Powering the /mem command in Telegram for instant personal knowledge retrieval.
- Searching through massive local documentation sets for specific headers or sections.
- Providing AI agents with a long-term memory cache of previous project decisions.
- Reducing latency in research workflows by bypassing cloud-based search APIs.
How Memory Search Works
- The user runs an indexing script to scan local files and build a optimized cached memory index.
- When a search is initiated, the system queries this local cache rather than the raw files.
- The search algorithm identifies the top matches based on the user's string or vector-based input.
- Results are returned with file paths, relevant headers, and brief summaries for immediate use.
Memory Search Setup
To begin using this tool within Openclaw Skills, you must first build your local index:
scripts/index-memory.py
Once the index is cached, you can perform searches by running:
scripts/search-memory.py "<your search query>" --top 5
Memory Search Data Schema & Taxonomy
| Component | Description |
|---|---|
| Cached Index | A local-first data structure stored on disk for fast retrieval. |
| Path Metadata | The absolute or relative file path to the source of the memory. |
| Headers | Section titles or organizational markers found within the indexed content. |
| Top Hits | The ranked list of the most relevant results for any given query. |
Memory Search Advanced Features
- Local-first architecture ensures all data remains private and accessible offline.
- Configurable result depth through the --top parameter for tailored search granularity.
- Seamless Telegram integration for triggering memory searches via bot commands.
- High-speed performance designed to work alongside other Openclaw Skills for agentic workflows.
SKILL.md
Loading
Related Openclaw Skills
gitbackup
Automatically generates self-contained Git bundles of your OpenClaw workspace to ensure your version history is backed up locally.
GitHub CLI (gh)
A powerful integration for managing GitHub repositories, issues, and pull requests via the official GitHub CLI.
Full Backup
A specialized skill for generating comprehensive local archives of the OpenClaw workspace and configuration files.
TrueProfit Profit Analytics Assistant
An advanced profit analytics skill for e-commerce merchants to track real-time net profit and advertising performance.
Search Memory
A high-performance local indexing and search utility designed to retrieve information from project memory files using keyword scoring and recency boosts.
Openclaw Starter Skill
A foundational template designed to integrate custom capabilities and automated workflows into the Openclaw environment.
Featured*
