Structured thinking, amplified

Choose from 14 dialog frameworks designed for different types of problems. Each framework guides how your AI agents interact—ensuring productive, focused conversations.

14 frameworks for every challenge

Adversarial & Positional

When you need ideas stress-tested

⚔️

Debate

Two or more sides defend opposing positions with opening statements, rebuttals, and closing arguments.

Best for: Testing ideas, understanding opposing views, strengthening arguments
🔴

Red Team / Blue Team

One side attacks an idea, strategy, or plan. The other defends and adapts under pressure.

Best for: Security analysis, strategy validation, finding weaknesses
🤝

Negotiation

Agents have competing interests and must reach a deal or compromise.

Best for: Deal-making scenarios, stakeholder alignment, conflict resolution

Collaborative & Knowledge-Building

When you need to solve together

🗳️

Deliberation

Agents share a problem and work collaboratively toward deciding on a course of action.

Best for: Team decisions, policy-making, strategic planning
🔍

Inquiry

Agents jointly investigate a question that none has a predetermined answer to.

Best for: Research, exploration, learning new domains
🧩

Problem Solving Pipeline

Structured phases: define problem → explore options → converge on solution.

Best for: Complex problems, systematic analysis, action planning

Perspective & Analysis

When you need comprehensive coverage

🎩

Six Thinking Hats

Each agent wears a different 'hat' (facts, emotions, caution, benefits, creativity, process).

Best for: Comprehensive analysis, avoiding blind spots, group decisions
🔄

Reverse Brainstorm

"How could we make this problem worse?" Then flip insights into solutions.

Best for: Innovation, risk identification, creative problem-solving
🎤

Interview / Interrogation

One agent deeply questions another. Can be Socratic, journalistic, or cross-examination style.

Best for: Deep exploration, uncovering assumptions, expertise extraction

Iterative & Loop-Based

When you need continuous improvement

🔁

Ralph Wiggum Loop

One agent builds solutions while another asks naive, challenging questions. Rebuild until bulletproof.

Best for: Stress-testing ideas, finding edge cases, robust solutions
🪞

Reflexion Loop

Agent produces output, self-critiques, generates lessons learned, and tries again with accumulated wisdom.

Best for: Quality improvement, learning from mistakes, iterative refinement
🎯

OODA Loop

Observe → Orient → Decide → Act → repeat. A military decision-making framework adapted for AI.

Best for: Strategy, competitive analysis, crisis response, scenario planning

Creative & Open

When you need to explore possibilities

Critique & Improve

One agent presents something, others critique it, then collaboratively improve it.

Best for: Content refinement, peer review, quality assurance
🎭

Simulation / Role-Play

Agents act out a scenario: boardroom meeting, negotiation, trial, pitch, historical event.

Best for: Training, scenario exploration, creative writing, preparation

Find the right framework for your challenge

Explore Frameworks