A technical framework to equip AI agents with long-term memory through persistent context storage and vector-based search.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install memory-setup-1-0-0
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 memory-setup-1-0-0 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 Setup?
The memory-setup skill is designed to solve the common goldfish brain problem in AI coding assistants like Moltbot and Clawdbot. By implementing this skill, developers can create a permanent record of decisions, preferences, and project history that the agent can reference across multiple sessions. It leverages vector search to retrieve relevant information from a structured file system, ensuring that your agent maintains continuity and understands the evolution of your codebase.
This implementation of Openclaw Skills bridges the gap between ephemeral chat sessions and a truly intelligent, context-aware partner. By centralizing long-term memory in Markdown files, the system remains transparent, editable by humans, and highly portable across different development environments.
Memory Setup Use Cases
- Retrieving architectural decisions made in previous development cycles.
- Maintaining a consistent record of user preferences and coding styles.
- Tracking project-specific context and status updates across long-term engagements.
- Accessing historical conversation transcripts to resume complex workflows.
- Managing daily logs to provide the agent with a chronological understanding of progress.
How Memory Setup Works
- The user enables the memorySearch module within the agent configuration file using a supported embedding provider.
- A standardized workspace directory structure is created to house long-term memory files and daily logs.
- The agent indexes these Markdown files using vector embeddings to create a searchable semantic index.
- During a conversation, the agent triggers memory_search tools to query the index for relevant context based on the current prompt.
- Retrieved snippets are injected into the agent's current context window to provide high-fidelity continuity and informed responses.
Memory Setup Setup
To begin using this skill within your Openclaw Skills workflow, update your ~/.clawdbot/clawdbot.json (or moltbot.json) configuration:
{
"memorySearch": {
"enabled": true,
"provider": "voyage",
"sources": ["memory", "sessions"],
"indexMode": "hot",
"minScore": 0.3,
"maxResults": 20
}
}
Initialize your workspace structure with the following commands:
mkdir -p memory/{logs,projects,groups,system}
touch MEMORY.md
Finally, restart the gateway to apply the configuration:
clawdbot gateway restart
Memory Setup Data Schema & Taxonomy
The skill organizes data into a hierarchical Markdown-based structure optimized for vector retrieval and human readability:
| File/Folder | Description | Metadata/Content |
|---|---|---|
MEMORY.md |
Core long-term memory | User facts, project summaries, and key decisions |
memory/logs/ |
Daily activity logs | YYYY-MM-DD.md files capturing task-specific events |
memory/projects/ |
Project context | Deep-dive documentation for specific repositories |
memory/system/ |
System preferences | Technical setup notes and tool-specific configurations |
Memory Setup Advanced Features
- Real-time indexing with hot mode to ensure newly created logs are searchable immediately.
- Multi-provider support allowing seamless switching between Voyage AI, OpenAI, and local embedding models.
- Customizable relevance thresholds via minScore to tune the precision and recall of retrieved memories.
- Agent behavior steering via AGENTS.md to define specific recall protocols and search triggers.
- Automated session history indexing to transform past conversation transcripts into searchable knowledge assets.
SKILL.md
Loading
Related Openclaw Skills
Copywriter
A precision copywriting tool designed to generate user-centric UX microcopy, high-conversion marketing text, and professional product messaging.
Auto-Updater
A background utility that automatically keeps your core bot and all installed Openclaw Skills updated to the latest versions daily.
Agent Browser
A fast, Rust-based headless browser CLI that allows AI agents to interact with web pages through simple, structured commands.
Protonmail
A powerful integration for AI agents to manage Protonmail accounts using Proton Bridge and the himalaya CLI client.
Xiaohongshu Title Generator
A specialized AI agent skill that generates high-conversion Xiaohongshu titles by leveraging emotional hooks, platform-specific algorithms, and native style mapping.
Built at GrowthX Project Submission
A specialized skill for GrowthX members to automatically discover and submit their development projects to the Built at GrowthX community showcase.
Featured*
