Knuckles-Team Paper Scan Workflow
CONCEPT:RESEARCH-001
Steps
Step 1: Topic Extractor
Agent: search-agent
Tools: sx_search, graph_query
Extract focus topics from the Knowledge Graph to build a relevance taxonomy.
Step 2: Scholarx Fetcher
Agent: analyzer-agent
Tools: graph_analyze, sx_storage
Fetch daily papers via the scholarx MCP using the extracted taxonomy.
Step 3: Paper Scorer
Agent: synthesizer-agent
Tools: graph_analyze, document_tools
Score the fetched papers locally against the relevance taxonomy.
Step 4: Paper Downloader
Agent: ingestor-agent
Tools: graph_write, kg_graph_ingest
Bulk download the most valuable papers for ingestion.
Step 5: KG Persistence [depends_on: paper-downloader]
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
- Paper Scan results persisted in KG
- Structured report (MD/PDF)
- Audit trail with timestamps and agent attributions
Execution
Run this workflow as a dependency-ordered DAG. Steps with no unmet depends_on run in parallel; dependents run after their prerequisites complete.
- Run first (in parallel): Step 1 — Topic Extractor; Step 2 — Scholarx Fetcher; Step 3 — Paper Scorer; Step 4 — Paper Downloader
- After level 0: Step 5 — KG Persistence
Execution: If graph-os is reachable, offload the whole DAG via graph_orchestrate action=execute_workflow (or the kg-delegate skill) for true parallel/swarm execution. Otherwise execute the steps natively in dependency order: run steps with no unmet depends_on in parallel, then their dependents.