What problem does it solve?
A lightweight, composable toolkit of reasoning primitives that helps agents and humans break decision paralysis, surface hidden assumptions, and stress-test plans before committing to action. It reduces costly mistakes by turning informal intuition into small, auditable thinking steps that an AI or a team can follow and reproduce.
Core Features & Use Cases
- Composable Primitives: Twelve self-contained protocols (assumption-audit, pre-mortem, steelman, matrix-evaluation, etc.) that can be run individually or chained into pipelines for full decision audits.
- Triage & Modes: A fast triage guide to pick the right primitive(s) for the situation and three explain modes (none / scaffold / full) to control verbosity and disclosure of intermediate artifacts.
- References & Operational Protocols: Detailed, short primitive guides and a triage reference that make it easy to inject structured thinking into subagent prompts for code reviews, product choices, estimations, and high-stakes decisions.
Quick Start
Ask the agent to run an assumption-audit and pre-mortem on your plan, then return a concise recommendation with top mitigations.