A comprehensive schema lifecycle management tool for organizing, validating, and evolving structured notes in AI-driven knowledge graphs.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install memory-schema
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-schema 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 Schema Management?
The Memory Schema skill provides a robust framework for managing structured note types using the Picoschema system. By defining clear schemas for entities like meetings, people, and tasks, users can ensure their personal knowledge management system remains uniform, queryable, and highly organized. This skill is essential for developers and power users utilizing Openclaw Skills to build complex, interconnected data structures where consistency is paramount.
At its core, Memory Schema allows an AI agent to infer structures from existing content, validate new entries against predefined rules, and detect drift as the knowledge base evolves. By utilizing YAML frontmatter and Picoschema syntax, it creates a bridge between unstructured markdown notes and a formal knowledge graph, making it a foundation for advanced data retrieval and agentic workflows.
Memory Schema Management Use Cases
- Standardizing emerging note types such as project decisions, research logs, or person profiles.
- Validating that existing notes conform to required fields and data types to prevent data corruption.
- Detecting schema drift when notes begin to include new fields that haven't been formalized yet.
- Managing the evolution of data structures through versioned schema updates as project requirements change.
- Automating the creation of structured metadata for use in broader Openclaw Skills ecosystems.
How Memory Schema Management Works
- Analyze existing note clusters to identify shared structures and automatically infer a candidate schema.
- Create a formal schema definition note within the schema directory, specifying required and optional fields.
- Implement Picoschema syntax to define data types such as strings, integers, enums, and arrays.
- Run validation routines to check note conformance and identify missing fields or type mismatches.
- Perform drift analysis to find discrepancies between the defined schema and actual note usage over time.
- Update and version schemas to accommodate new data patterns while maintaining backward compatibility.
Memory Schema Management Setup
To set up a new schema within the Openclaw Skills framework, create a markdown file in your schema directory. You can use the write_note tool to initialize the definition.
# Example of creating a schema via the agent
write_note(
title="Meeting",
directory="schema",
note_type="schema",
metadata={
"entity": "Meeting",
"version": 1,
"schema": {
"topic": "string",
"date": "string",
"status(enum)": ["scheduled", "completed"]
}
}
)
Ensure your agent has the necessary permissions to read from and write to the schema/ directory to manage the lifecycle effectively.
Memory Schema Management Data Schema & Taxonomy
Memory Schema organizes its metadata using a clear taxonomy within the YAML frontmatter. This structure is critical for integration with other Openclaw Skills.
| Property | Type | Description |
|---|---|---|
entity |
String | The name of the note type being defined (e.g., Task, Person). |
version |
Integer | The schema version, incremented during breaking changes. |
schema |
Object | The Picoschema definition mapping fields to types and descriptions. |
settings |
Object | Configuration for validation strictness (e.g., warn or error). |
relations |
Type | Direct references to other entity types to build graph edges. |
Memory Schema Management Advanced Features
- Automated Schema Inference: Generate suggested schemas by analyzing the threshold of field frequency across note sets.
- Relational Knowledge Mapping: Define edges between entities (e.g., linking a 'Meeting' to 'Person' attendees) directly in the schema.
- Multi-type Validation: Support for complex arrays and enums to ensure high-fidelity data entry.
- Drift Detection: Identify fields used in notes that are missing from the schema, ensuring Openclaw Skills stay updated with user habits.
- Flexible Validation Modes: Switch between 'warn' for organic growth and 'error' for strict data pipelines.
SKILL.md
Loading
Related Openclaw Skills
Article Rewriter
A high-performance content optimization engine that uses the APAG framework to turn rough drafts into persuasive, polished articles.
Calculator
A robust arithmetic engine for performing deterministic mathematical calculations and multi-step expression evaluation.
PDFExtract Pull Text from PDFs
A local extraction utility that converts complex PDF documents into clean, structured markdown for AI agents without external services.
HiNiHao Chinese Tutor
HiNiHao is a proactive AI Mandarin tutor that delivers personalized Chinese lessons using authentic media and real-world content on a scheduled basis.
Shopify Toolkit
A comprehensive development reference for building Shopify themes, apps, and headless storefronts using Liquid and GraphQL.
Maine Coon Virtual Pet
A sophisticated virtual pet skill for AI agents to adopt and raise a gentle giant Maine Coon through real-time resource management.
Featured*
