Model and store typed knowledge graphs with schema constraints and append-only logging.
npx skills add https://github.com/syahravi/openclaw --skill ontology-syahravi
Please help me install this Agent Skill. Skill: ontology Source: https://github.com/syahravi/openclaw/tree/main/skills/ontology Command: npx skills add https://github.com/syahravi/openclaw --skill ontology-syahravi
Many agent workflows suffer from fragmented, inconsistent, or unstructured memory that makes linking people, projects, tasks, events, and documents difficult to query or validate. The ontology skill provides a typed knowledge graph with relations and constraints so agents and skills can reliably store, discover, and reason about shared state.
Create a Person named Alice and a Project called Website Redesign, then relate the project to Alice as its owner using the ontology CLI.
High-intent search queries and answers about installing and using this skill.
Structured agent memory is stored by modeling typed entities and relations in a local knowledge graph, enforcing schema constraints and append-only mutation logging to ensure data consistency and safe merges.
A typed knowledge graph validates relations by enforcing schema constraints, required properties, enums, and relation cardinality, plus acyclicity checks to prevent circular dependencies during entity linking and CRUD operations.
Multi-step planning is modeled as graph transformations by creating validated create and relate operations for entities, integrating cross-skill communication patterns for shared state within local workspace workflows.
You can evolve entity types using schema-append utilities that merge new properties without overwriting past data, maintaining immutable history through append-only graph storage for safe migrations.
Entities are created and linked using a CLI-driven CRUD interface that supports relation creation, graph queries, traversal utilities, and a programmatic API for integration with other skills.
A typed knowledge graph is needed when workflows suffer from fragmented memory, requiring reliable storage, discovery, and reasoning about shared state across people, projects, tasks, events, and documents.
MaosixianCreates typed knowledge graphs for agent memory and shared state using Python and JSONL files.
getspurQuery a code repository's graph artifact using DuckDB with DuckPGQ and Onager.
daffy0208Design and maintain knowledge bases with RAG, graph, and hybrid architectures.
bobvarkeyStore and validate structured knowledge as a typed entity graph.
