Documentation To Kg Workflow
CONCEPT:RESEARCH-001
Parallel execution workflow for documentation to kg using the Unified Parallel Engine
Steps
Step 1: Parse Markdown
Agent: search-agent
Tools: sx_search, graph_query
Execute parse markdown operations for the Documentation To Kg workflow.
Expected: parse_markdown_artifacts
Step 2: Extract Concepts [depends_on: parse_markdown]
Agent: analyzer-agent
Tools: graph_analyze, sx_storage
Execute extract concepts operations for the Documentation To Kg workflow.
Expected: extract_concepts_artifacts
Step 3: Entity Resolution [depends_on: extract_concepts]
Agent: synthesizer-agent
Tools: graph_analyze, document_tools
Execute entity resolution operations for the Documentation To Kg workflow.
Expected: entity_resolution_artifacts
Step 4: Ingest [depends_on: entity_resolution]
Agent: ingestor-agent
Tools: graph_write, kg_graph_ingest
Execute ingest operations for the Documentation To Kg workflow.
Expected: ingest_artifacts
Step 5: KG Persistence [depends_on: ingest]
Agent: ingestor-agent
Tools: graph_write
Persist workflow results as nodes and edges in the Knowledge Graph. Create appropriate typed nodes with metadata and link to existing domain entities.
Output
- Documentation To Kg results persisted in KG
- Structured report (MD/PDF)
- Audit trail with timestamps and agent attributions
Source: Knuckles-Team/universal-skills — distributed by TomeVault.