What problem does it solve?
This Skill helps you convert messy trading inputs (market context, trade logs, and journal snippets) into structured, falsifiable trade hypotheses you can actually test.
Core Features & Use Cases
- Evidence-driven hypothesis generation: Produces 1–5 hypothesis cards by extracting winning/losing patterns and constraints from a normalized input bundle.
- Critique and ranking with guardrails: Applies a critique pass, then ranks hypotheses into prioritized cards with explicit kill criteria and minimum viable experiments.
- Optional strategy export (v1): Exports “pursue” hypotheses into
strategy.yaml + metadata.json artifacts compatible with edge-finder-candidate/v1.
- Use Case: When you refine an earnings setup (e.g., VWAP hold or breakout timing) you can generate hypothesis experiments that focus on objective invalidation rules rather than intuition.
Quick Start
Run pass 2 by providing a normalized input bundle JSON and a raw hypotheses JSON, then write ranked output artifacts to your chosen reports directory.