# Overnight Research

> Unattended overnight research loop: Generate → Backtest → Debate → Report.

- Skill: `knuckles-team/overnight-research` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add knuckles-team/overnight-research`
- Raw SKILL.md: https://api.skillmd.com/api/skills/knuckles-team/overnight-research/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Docs & Writing
- Author: Knuckles-Team (https://skillmd.com/u/knuckles-team)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/knuckles-team/overnight-research

---

# Overnight Research Workflow (Cron: 10PM ET daily)

## Workflow Execution Steps

### Step 1: hypothesis-generation
Generate N hypotheses from KG signals and market data.
Tool: `emerald_signals(action="alpha")` + KG research query.

### Step 2: batch-backtest
Run all hypotheses through qlib backtest queue on GPU node.
Tool: Route to `data-science-mcp` for batch execution.

### Step 3: debate-winners
Debate top-performing hypotheses via trading swarm.
Tool: `graph_orchestrate(action="start_debate", task=top_hypothesis)`

### Step 4: kg-persist
Store validated strategies and research results in KG.
Tool: `graph_write(action="add_node", node_type="BacktestResult", ...)`

### Step 5: report
Generate research report with rankings and recommendations.

## 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 — hypothesis-generation; Step 2 — batch-backtest; Step 3 — debate-winners; Step 4 — kg-persist; Step 5 — report

**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.