By name: market_microstructure_analysis description: >- Audits millisecond order book snapshot logs, computes VPIN metrics, decomposes spreads, and aggregates impact reports. domain: finance agent: quant_analyst team_config: name: quantitative_trading_team task_pattern: quantitative analysis and financial computation execution_mode: parallel specialist_ids: - data-fetcher - compute-engine - risk-assessor - report-generator tool_assignments: data-fetcher: [graph_query, sx_search] compute-engine: [graph_analyze] risk-assessor: [graph_query, graph_analyze] report-generator: [graph_write, document_tools] tags: [microstructure, order-book, vpin, slippage]
concept: CONCEPT:EE-011
Market Microstructure Analysis Workflow
CONCEPT:EE-011
Audits millisecond order book snapshot logs, computes VPIN metrics, decomposes spreads, and aggregates impact reports.
Steps
Step 1: Book Snapshot Crawlers
Agent: data-fetcher
Tools: graph_query, sx_search
Collects millisecond order book snapshot logs.
Expected: order-book-snapshots
Step 2: Flow Toxicity Calculator [depends_on: book-snapshot-crawlers]
Agent: compute-engine
Tools: graph_analyze
Calculates VPIN and buy-sell order flow imbalance.
Expected: volume-toxicity-signals
Step 3: Spread Decomposition [depends_on: flow-toxicity-calculator]
Agent: risk-assessor
Tools: graph_query, graph_analyze
Decomposes bid-ask spread into inventory and adverse selection.
Expected: decomposed-spread-metrics
Step 4: Impact Report [depends_on: spread-decomposition]
Agent: report-generator
Tools: graph_write, document_tools
Compiles execution cost optimization recommendation cards.
Expected: microstructure-impact-tearsheet
Step 5: KG Persistence [depends_on: impact-report]
Agent: report-generator
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
- Market Microstructure Analysis 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 — Book Snapshot Crawlers
- After level 0: Step 2 — Flow Toxicity Calculator
- After level 1: Step 3 — Spread Decomposition
- After level 2: Step 4 — Impact Report
- After level 3: Step 5 — KG Persistence
Execution: If graph-os is reachable, offload the whole DAG via graph_orchestrate action=execute_workflow (or the kg-delegation-router 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.