A local-first knowledge layer for AI agents that validates patterns through repeated successful use to ensure high-reliability memory.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install memory-garden
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-garden 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 Garden?
Memory Garden is a sophisticated knowledge management skill designed to solve the problem of AI agents forgetting what worked in previous sessions. By implementing a unique N-count convergence logic, it ensures that only patterns that have proven themselves through repeated use are treated as validated knowledge. This local-first system allows users to curate a private library of domain-specific wisdom that persists across projects, making it a cornerstone for those building robust workflows with Openclaw Skills.
The skill operates as a background daemon that manages pattern extraction, search, and validation. It distinguishes itself from simple vector databases by requiring evidence of success before a pattern is promoted. This ensures your agent is guided by reliable, proven advice rather than one-off observations, significantly improving the quality of long-term agent interactions.
Memory Garden Use Cases
- Reducing repetitive errors by allowing agents to surface previously successful solutions.
- Building a local repository of project-specific coding patterns and architectural decisions.
- Sharing trusted knowledge across distributed teams using peer-to-peer synchronization.
- Auditing the effectiveness of AI-generated fixes through category-based success tracking.
How Memory Garden Works
- The agent identifies a recurring successful workflow and records it as an initial observation (N=1).
- Subsequent successful applications of the same pattern increment the N-count, confirming its recurring utility (N=2).
- Once a pattern reaches a validation threshold (N=3+), it is prioritized as high-trust knowledge for future queries.
- The agent utilizes the search_patterns tool to augment its context with these validated insights during new tasks.
- Optional P2P synchronization allows these patterns to be shared across a fleet of agents or different machines.
Memory Garden Setup
To get started with this skill, ensure the daemon is installed and properly configured within your environment. This is a vital addition to your collection of Openclaw Skills.
# Install the skill and its dependencies
clawhub install memory-garden
# Verify the installation and daemon status
clawhub inspect memory-garden
# Perform a health check on the local daemon
curl http://127.0.0.1:18790/health
Memory Garden Data Schema & Taxonomy
Memory Garden prioritizes data sovereignty, keeping all patterns and logs on your local filesystem unless you explicitly enable synchronization.
| Data Type | Storage Location | Description |
|---|---|---|
| Pattern Store | ~/.memory-garden/ |
Local database containing extracted patterns and their N-count metadata. |
| Operation Logs | ~/.memory-garden/logs/ |
Detailed logs for troubleshooting daemon activity and tool calls. |
| Environment Config | Shell Profile | Configuration for features like MG_EXTRACTION_ENABLED and MG_SYNC_ENABLED. |
Memory Garden Advanced Features
- Federated curation: Subscribe to trusted pattern sources while maintaining local control.
- Continuous Improver Integration: Use webhooks to track which patterns successfully pass CI/CD checks.
- Opt-in P2P Sync: Share validated knowledge across your local network without cloud intermediaries.
- Human-in-the-loop Extraction: Ensure only high-quality patterns are planted by requiring manual confirmation.
SKILL.md
Loading
Related Openclaw Skills
insight-song
A creative tool that synthesizes technical conversations into professionally formatted song lyrics and style tags for Suno.ai.
failure-memory
failure-memory is a specialized system for Openclaw Skills that enables AI agents to detect, record, and learn from their own failures to prevent future recurrences.
Ethics Guardrails
A public API-backed framework for declaring and publishing an AI agent's ethical boundaries and priority hierarchies.
context-verifier
A foundational verification system that ensures file integrity and detects data drift before an AI agent performs operations.
Moltbook Search
Perform advanced hybrid semantic search across a massive database of over 125,000 AI agent posts from the Moltbook social network.
Patent Scanner
A technical analysis tool that identifies distinctive patterns in concepts using a structured patent attorney methodology.
Featured*
