# Strategy Tournament

> Performs concurrent backtests across 10 distinct algorithmic strategies, evaluates risk-adjusted metrics, ranks return profiles, and selects the top models.

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

---


# Strategy Tournament Workflow

**CONCEPT:EE-011**

Performs concurrent backtests across 10 distinct algorithmic strategies, evaluates risk-adjusted metrics, ranks return profiles, and selects the top models.

## Steps

### Step 1: Backtest Batch Runner
**Agent**: `data-fetcher`
**Tools**: `graph_query, sx_search`

Runs historical backtests concurrently for 10 distinct trading strategies.
Expected: `batch-backtest-outputs`

### Step 2: Metric Evaluator [depends_on: backtest-batch-runner]
**Agent**: `compute-engine`
**Tools**: `graph_analyze`

Computes Sharpe, Sortino, Calmar ratios, and max drawdown for each candidate.
Expected: `strategy-performance-scorecards`

### Step 3: Ranker Selector [depends_on: metric-evaluator]
**Agent**: `risk-assessor`
**Tools**: `graph_query, graph_analyze`

Ranks strategies and selects the top 3 best-performing models.
Expected: `winning-strategies-selection`

### Step 4: Synthesis Compiler [depends_on: ranker-selector]
**Agent**: `report-generator`
**Tools**: `graph_write, document_tools`

Synthesizes results into a unified strategic tearsheet.
Expected: `tournament-summary-report`

### Step 5: KG Persistence [depends_on: synthesis-compiler]
**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
- Strategy Tournament 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 — Backtest Batch Runner
- **After level 0:** Step 2 — Metric Evaluator
- **After level 1:** Step 3 — Ranker Selector
- **After level 2:** Step 4 — Synthesis Compiler
- **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-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.