← Back to Skills Marketplace
Agent Ontology
by
Fang, Chao
· GitHub ↗
· v1.0.0
· MIT-0
110
Downloads
0
Stars
1
Active Installs
1
Versions
Install in OpenClaw
/install agent-ontology
Description
Create, update, and query typed entities and relations like Person, Project, and Task in a verifiable knowledge graph with validation.
README (SKILL.md)
Ontology Skill
Core Concept
A typed vocabulary + constraint system for representing knowledge as a verifiable graph.
When to Use
- "Remember that..." → Create/update entity
- "What do I know about X?" → Query graph
- "Link X to Y" → Create relation
- "Show all tasks for project Z" → Graph traversal
Storage
memory/ontology/graph.jsonl - append-only JSONL format
Core Types
- Person: name (required), email?, phone?, notes?
- Project: name (required), status?, goals[], description?
- Task: title (required), status (required), due?, priority?, assignee?
Relations
| Relation | From | To |
|---|---|---|
| has_owner | Project, Task | Person |
| assigned_to | Task | Person |
| belongs_to | Task | Project |
| blocks | Task | Task (acyclic) |
CLI Commands
python3 scripts/ontology.py create --type Person --props '{"name":"Alice"}'
python3 scripts/ontology.py list --type Person
python3 scripts/ontology.py query --type Task --where '{"status":"open"}'
python3 scripts/ontology.py get --id p_001
python3 scripts/ontology.py relate --from proj_001 --rel has_owner --to p_001
Validation
- Required properties must exist
- Enum values from predefined sets
- No circular dependencies for acyclic relations
Usage Guidance
Do not treat this as a complete approval; rerun the review in an environment where metadata.json and artifact/ can be read.
Capability Tags
Capability Assessment
Purpose & Capability
Workspace command execution failed before files could be read, so purpose and capabilities could not be verified from artifact evidence.
Instruction Scope
No SKILL.md content was available for scope review due the workspace inspection failure.
Install Mechanism
No install metadata or artifact manifest content was available for review.
Credentials
Environment access could not be assessed because artifact contents were not readable through the available tool session.
Persistence & Privilege
No artifact-backed persistence or privilege concern was identified, but the relevant files could not be inspected.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install agent-ontology - After installation, invoke the skill by name or use
/agent-ontology - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
- Initial release of agent-ontology skill.
- Provides a typed vocabulary and constraint system for knowledge graphs.
- Supports core entity types: Person, Project, and Task.
- Defines key relations between entities, including ownership, assignment, and task dependencies.
- Stores data in an append-only JSONL file at memory/ontology/graph.jsonl.
- Offers CLI commands for creating, querying, and relating entities.
- Includes basic validation for required fields, enums, and acyclic relations.
Metadata
Frequently Asked Questions
What is Agent Ontology?
Create, update, and query typed entities and relations like Person, Project, and Task in a verifiable knowledge graph with validation. It is an AI Agent Skill for Claude Code / OpenClaw, with 110 downloads so far.
How do I install Agent Ontology?
Run "/install agent-ontology" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Agent Ontology free?
Yes, Agent Ontology is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Agent Ontology support?
Agent Ontology is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Agent Ontology?
It is built and maintained by Fang, Chao (@biociao); the current version is v1.0.0.
More Skills