A typed vocabulary and constraint system for representing agent knowledge as a verifiable, structured graph.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install zqtest
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 zqtest 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 Ontology?
The Ontology skill provides a robust framework for managing an AI agent's memory by organizing information into a typed knowledge graph. Instead of relying on unstructured text, this skill allows agents to define entities such as people, projects, and tasks with specific properties and relationships. By using Openclaw Skills with a formal ontology, developers can ensure data integrity through schema validation and complex graph traversals.
This system acts as a central source of truth, enabling agents to remember context across different sessions and reason about the world through linked data. Whether tracking project dependencies or mapping organizational structures, the Ontology skill provides the necessary scaffolding for sophisticated agentic workflows.
Ontology Use Cases
- Building a persistent memory of people, projects, and organizational structures.
- Managing complex task dependencies and identifying workflow blockers.
- Sharing structured state between multiple specialized AI agents.
- Validating input data against strict schema constraints to prevent hallucinations.
- Transforming high-level plans into a sequence of verifiable graph operations.
How Ontology Works
- Initialize the ontology environment by creating the storage files and optional schema definitions.
- Define entities by assigning them a specific type, such as a Task or a Document, along with required properties.
- Establish relationships between entities to create a network of interconnected data points.
- Perform graph queries to retrieve specific entities or traverse links to find related information.
- Apply validation rules to ensure all mutations respect the defined schema and logic constraints like acyclicity.
Ontology Setup
To begin using this skill within your Openclaw Skills environment, initialize the storage directory and create the base graph file:
mkdir -p memory/ontology
touch memory/ontology/graph.jsonl
You can then optionally define a schema to enforce data types:
python3 scripts/ontology.py schema-append --data '{
"types": {
"Task": { "required": ["title", "status"] },
"Person": { "required": ["name"] }
}
}'
Ontology Data Schema & Taxonomy
The skill utilizes an append-only storage model to preserve history and ensure data reliability. Data is primarily stored in memory/ontology/graph.jsonl and memory/ontology/schema.yaml.
| Object | Description |
|---|---|
| Entity | A core data object containing a unique ID, Type, and key-value properties. |
| Relation | A link between two entities (from_id to to_id) with a specific relation type. |
| Constraint | Schema-level rules defining required fields, enums, and relation cardinality. |
| Operation | Log entries for create, relate, and update actions recorded in the JSONL file. |
Ontology Advanced Features
- Planning as Graph Transformation: Model multi-step agent plans as atomic changes to the knowledge graph.
- Acyclic Dependency Enforcement: Automatically prevent circular logic in task management or organizational hierarchies.
- Cross-Skill Interoperability: Use the ontology as a shared data layer for different Openclaw Skills to communicate.
- Causal Inference Integration: Log memory mutations as causal actions to track the history and rationale of agent decisions.
- Automatic Validation: Enforce logical constraints such as ensuring event end dates occur after start dates.
SKILL.md
Loading
Related Openclaw Skills
Daily Literature Search
An automated system for researchers to perform scheduled literature searches across multiple databases with integrated deduplication and PDF management.
Meta Ads Creator
An automated workflow for programmatically creating and managing Meta ad campaigns through the Marketing API.
SmartPi IoT Control
A dedicated Openclaw Skills plugin for controlling and monitoring SmartPi-enabled IoT devices through a standardized API.
Content Claw
Content Claw is a sophisticated content generation engine that transforms research papers, podcasts, and URLs into platform-specific social media content and visuals.
aicoin-monitor
A fully independent and customizable cryptocurrency data monitoring skill for tracking over 30 market metrics including institutional flows and on-chain health.
Zilliz Cloud CLI Skill
A powerful Openclaw Skills integration to manage Zilliz Cloud clusters and Milvus vector databases through natural language commands.
Featured*
