Research Intelligence Pipeline (CONCEPT:AU-KG.research.research-pipeline-runner)¶
Overview¶
Automated end-to-end research ingestion: ScholarX Discovery → 9-domain Relevance Scoring → Tiered Ingestion (full for ≥3.0, abstract-only for ≥1.0) → OWL Enrichment → Digest Generation. Supports arXiv, local files, and web URLs.
Implementation Details¶
- Source Code:
agent_utilities/automation/research_pipeline.py - Pillar: KG
Both relevance tiers use one authoritative graph-write contract. The Article,
Source, pseudonymous author references, and their edges are committed as one
engine-native ChangeEnvelope graph slice; there is no direct
engine.graph/backend upsert or ScholarX bridge fallback. Raw author names and
emails are not durable metadata: only stable non-reversible references cross the
persistence boundary.
When a full paper has a PDF, the article slice commits first. Full text is then
read from the local file without persisting its path, scrubbed of known author
identifiers, and materialized as a Document/Chunk slice through
DocumentProcessor's native ChangeEnvelope path. That second slice is an
optional retrieval projection, not a competing graph authority; a projection
failure is logged without undoing the already committed Article.
Documentation Coverage¶
This is an auto-generated dedicated concept page to ensure 100% documentation coverage across the ecosystem.
KG Source Resolver (CONCEPT:AU-KG.research.research-pipeline-runner)¶
Overview¶
Bridges the KG indexing layer to the comparative-analysis skill by materializing stored documents to filesystem paths with metadata enrichment. Optional — gracefully returns empty when no KG is available.
Implementation Details¶
- Source Code:
agent_utilities/knowledge_graph/core/source_resolver.py - Pillar: KG
Documentation Coverage¶
This is an auto-generated dedicated concept page to ensure 100% documentation coverage across the ecosystem.
Research Intelligence Sub-Agent (CONCEPT:AU-KG.research.research-pipeline-runner)¶
Overview¶
Isolated research context with citation graph traversal (Semantic Scholar API), doom-loop detection, and KG persistence. Findings become EvidenceNode entries with wasDerivedFrom provenance chains. Adapted from ml-intern's research_tool.py sub-agent pattern.
Implementation Details¶
- Source Code:
agent_utilities/knowledge_graph/orchestration/research_subagent.py - Pillar: KG
Documentation Coverage¶
This is an auto-generated dedicated concept page to ensure 100% documentation coverage across the ecosystem.
Research Orchestration Integration (CONCEPT:AU-KG.research.research-pipeline-runner)¶
Overview¶
Connects ResearchSubagent (AU-KG.research.zero-llm-pack-link) to ResearchPipelineRunner (KG-2.3) and UnifiedRAGKGRetriever (KG-2.38) for automated daily research cycles. 7-phase pipeline: discovery → subagent session → citation traversal → pipeline ingestion → similarity linking → cluster refresh → KG persistence. MCP-compatible for run_research_cycle tool registration.
Implementation Details¶
- Source Code:
agent_utilities/knowledge_graph/orchestration/research_orchestrator.py - Pillar: KG
Documentation Coverage¶
This is an auto-generated dedicated concept page to ensure 100% documentation coverage across the ecosystem.