Nous.KnowledgeBase.Workflows (nous v0.16.0)
View SourcePre-built workflow DAG pipelines for knowledge base operations.
Provides ready-made workflows for:
- Ingest pipeline — raw documents → compiled wiki entries
- Incremental update — detect changes and recompile affected entries
- Health check — audit the knowledge base for issues
- Output generation — produce reports, summaries, or slides
Summary
Functions
Builds a health check pipeline.
Builds an incremental update pipeline.
Builds an ingest pipeline workflow.
Builds an output generation pipeline.
Functions
Builds a health check pipeline.
Required input data
%{kb_config: %{store: ..., store_state: ..., kb_id: ...}}Builds an incremental update pipeline.
Detects changed documents and recompiles only affected entries.
Required input data
%{
documents: [%{title: "...", content: "..."}],
kb_config: %{store: ..., store_state: ..., kb_id: ...}
}Builds an ingest pipeline workflow.
Nodes: ingest_docs → extract_concepts → compile_entries → generate_links → embed_entries → persist
Required input data
%{
documents: [%{title: "...", content: "...", doc_type: :markdown}],
kb_config: %{store: ..., store_state: ..., kb_id: ...}
}Options
:compiler_model- Model string for LLM steps (default: "openai:gpt-4o-mini"):embedding- Embedding provider module:embedding_opts- Embedding options
Builds an output generation pipeline.
Required input data
%{
topic: "...",
output_type: :report | :summary | :slides,
kb_config: %{store: ..., store_state: ..., kb_id: ...}
}