A persistent long-term memory system that enables AI agents to store and retrieve context, preferences, and facts across multiple sessions.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install claw-memory-augment
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 claw-memory-augment 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 Augment?
Memory Augment is a sophisticated long-term memory layer designed specifically for the OpenClaw ecosystem. It bridges the gap between ephemeral chat sessions by providing a permanent storage solution for user preferences, technical decisions, and learned facts. By utilizing Openclaw Skills like this one, developers can ensure their agents maintain a consistent personality and deep understanding of project context over time.
This skill leverages hybrid search mechanics—combining keyword matching with semantic understanding—to retrieve the most relevant information exactly when the agent needs it. It transforms a standard AI into a truly personalized assistant that remembers past interactions and evolves based on user feedback, making it a cornerstone for complex, long-running development projects.
Memory Augment Use Cases
- Persisting user coding preferences and workspace configurations across different sessions.
- Tracking project decisions and approvals to prevent repetitive questioning.
- Maintaining a searchable history of learned technical patterns and facts.
- Automatically injecting relevant context into agent prompts based on the current task.
- Archiving session notes and task progress for long-term project management.
How Memory Augment Works
- The agent captures information during a session, such as a coding preference or a specific project decision.
- Information is stored in a structured storage layer (YAML or JSON) with metadata including tags, timestamps, and expiry dates.
- When a new query is initiated, the system performs a hybrid search to find relevant past memories using both keywords and semantic relevance.
- A scoring algorithm evaluates results based on recency, tag matches, and memory type (e.g., preferences are weighted higher than general context).
- Relevant memories are automatically injected into the agent context window before the next turn to provide instant awareness.
- The system learns from manual corrections logged by the user to improve search accuracy and ranking over time.
Memory Augment Setup
To get started with this addition to your Openclaw Skills, install it via the clawhub CLI:
npx clawhub install memory-augment
You can then configure your storage preferences and auto-injection rules in the ~/.memory-augment/config.yaml file.
Memory Augment Data Schema & Taxonomy
The skill manages data primarily through a structured YAML storage file located at ~/.memory-augment/storage.yaml.
| Field | Type | Description |
|---|---|---|
| id | UUID | Unique identifier for the memory entry. |
| content | String | The actual data, fact, or preference being remembered. |
| type | String | Categorization: preference, decision, context, or learning. |
| tags | Array | Custom keywords for organization and filtering. |
| score | Float | Confidence or relevance weight (0.0 to 1.0). |
| created | ISO8601 | Timestamp of when the memory was first recorded. |
| expires | ISO8601 | Optional expiration date for temporary session notes. |
Memory Augment Advanced Features
- Automatic context injection hooks triggered by specific keywords or intervals.
- Hybrid retrieval system combining semantic embeddings with keyword filtering.
- Memory decay logic that prioritizes high-impact information while archiving stale data.
- Integration capabilities with other Openclaw Skills like Cron Manager for automated summaries.
- Reinforcement learning via a corrections log to tune the scoring algorithm based on user feedback.
SKILL.md
Loading
Related Openclaw Skills
WPS/Word Automation
A Python-based automation utility for programmatically controlling Microsoft Word and WPS Writer on Windows systems.
AI Wallpaper Skill
A desktop enhancement module that generates high-quality AI images from prompts and automatically updates your system wallpaper.
Blogger Auto-Publish
A professional automation tool for publishing Markdown articles directly to Google Blogger with full OAuth 2.0 support and batch processing capabilities.
Research Daily Push
An automated research assistant that tracks the latest arXiv publications and delivers structured summaries to your preferred communication tools.
Challenge Loop
An adversarial hardening skill that uses self-refutation and independent subagent challenges to stress-test AI judgments and recommendations.
Inbox Triage
A powerful AI automation tool designed to filter, prioritize, and draft responses for overwhelming message volumes across multiple platforms.
Featured*
