How abhinavbwj/aec-knowledge-graph works
AEC Knowledge Graph turns curated JSONL entities and relationships, plus extracted skill references, into an embedded Kuzu graph that engineers can query from the CLI or Python. The main workflow is seed or plugin input → ETL/validation → database build → canonical Cypher queries → results. Query capabilities include influence paths, project discovery and profiles, movement and standards lineage, skill citations, and graph statistics. The source excerpts show the main wiring; the documented sibling-plugin inputs are external to this repository.
Main entry points: query.py Canonical queries · cli.py CLI · load.py Graph builder · from_trinity.py Skill ETL
Components 17
Graph inputs
- Seed entities
data/nodes - Seed relations
data/edges - ETL output
data/auto
Ingestion and storage
- Skill ETL
src/aec_kg/etl/from_trinity.py - Pydantic models
src/aec_kg/schema.py - Graph builder
src/aec_kg/load.py - Kuzu graph
src/aec_kg/load.py
Query and access
- CLI
src/aec_kg/cli.py - Canonical queries
src/aec_kg/query.py - Project discovery
src/aec_kg/query.py - Influence and lineage
src/aec_kg/query.py - Skill citations
src/aec_kg/query.py - Graph statistics
src/aec_kg/query.py - Query results
src/aec_kg/query.py
Other components
- Engineer
- Python callers
- Sibling plugins
Connections 28
- Engineer → CLI: invokes
src/aec_kg/cli.py - CLI → Skill ETL: runs ETL
src/aec_kg/cli.py - Sibling plugins → Skill ETL: provides skills
src/aec_kg/etl/from_trinity.py - Seed entities → Skill ETL: supplies aliases
src/aec_kg/etl/from_trinity.py - Skill ETL → Pydantic models: validates output
src/aec_kg/etl/from_trinity.py - Skill ETL → ETL output: writes JSONL
src/aec_kg/etl/from_trinity.py - CLI → Graph builder: builds graph
src/aec_kg/cli.py - Seed entities → Graph builder: loads entities
src/aec_kg/load.py - Seed relations → Graph builder: loads relations
src/aec_kg/load.py - ETL output → Graph builder: merges ETL data
src/aec_kg/load.py - Graph builder → Pydantic models: validates records
src/aec_kg/load.py - Graph builder → Kuzu graph: creates and populates
src/aec_kg/load.py - Engineer → Python callers: calls
README.md - Python callers → Canonical queries: requests queries
README.md - CLI → Canonical queries: dispatches queries
src/aec_kg/cli.py - CLI → Graph statistics: requests counts
src/aec_kg/cli.py - Canonical queries → Kuzu graph: opens connection
src/aec_kg/query.py - Canonical queries → Pydantic models: uses graph labels
src/aec_kg/query.py - Canonical queries → Project discovery: supports
README.md - Canonical queries → Influence and lineage: supports
README.md - Canonical queries → Skill citations: supports
README.md - Canonical queries → Graph statistics: counts labels
src/aec_kg/query.py - Project discovery → Query results: returns project data
src/aec_kg/query.py - Influence and lineage → Query results: returns paths
src/aec_kg/query.py - Skill citations → Query results: returns citations
src/aec_kg/query.py - Graph statistics → Query results: returns counts
src/aec_kg/query.py - Canonical queries → Query results: returns rows
src/aec_kg/query.py - CLI → Query results: prints results
src/aec_kg/cli.py