9.2
/ 10
1 evaluations
96.1k Downloads
Overview
Implements a local, file-based typed knowledge graph (ontology) with entity CRUD, relations, validation against type constraints, and append-only persistence. Supports Person, Project, Task, Event, Document, and other core types with schema-driven constraints.
Key Advantages
1.Structured entity-relation model with validation
2.Append-only graph storage preserves history
3.Schema in YAML for required fields, enums, acyclicity
4.CLI and script integration (scripts/ontology.py)
5.No secrets stored inline, workspace-scoped path resolver
Use Cases
- Remember that... / What do I know about X?
- Link X to Y / Show all tasks for project Z
- Planning multi-step work as graph transformations
- Cross-skill shared state (e.g. Commitment → Task)
- Causal inference action logging
Evaluation Scores
9.2
/ 10
Reliability
9.2
Functionality
9.4
Usability
8.9
Safety
9.5
Performance
8.8
Compatibility
9.0
Based on 1 evaluation · Latest: 3/18/2026
Download Trend
Loading...
Evaluation History (1)
9.2/103/18/2026▼
OS: darwin 25.3.0LLM: default
Ontology skill is internally consistent: local file-based graph (memory/ontology/graph.jsonl), no extra credentials or unusual installs. OpenClaw scan: Benign, high confidence. Purpose matches SKILL.md and Python script; runtime scope is explicit (local files only). Recommended to review scripts/ontology.py and validate_graph implementation before production use.
Comments (0)
No comments yet. Be the first!