The fastest way to make a worse decision is to only argue your own side. This tool takes any proposition and builds the strongest honest case both for and against it using your own OpenAI or Anthropic key — then tells you where the disagreement really turns, all in your browser.
How it works
Pick a provider and model, paste your API key, and choose a depth. Enter your proposition and any context about your situation. The tool builds a prompt that asks the model to steel-man — not straw-man — both sides, present them under clear “For” and “Against” headings, take no position, and close with a “Key cruxes” section naming the questions the debate actually hinges on. It makes one direct request to the provider and returns the full breakdown to copy.
For Anthropic, the request includes the official direct-browser-access header so it works straight from the page.
Where it helps most
This is useful any time you catch yourself already certain. Before a contentious decision, generate both cases and read the side you disagree with first — that is where the blind spots hide. For debate or interview prep it gives you the opposition’s best material so nothing on the day surprises you. The cruxes section is the real payoff: most arguments that feel intractable actually turn on one or two factual or values questions, and naming them turns an endless back-and-forth into something you can resolve.
Notes and limits
The model argues persuasively for positions regardless of whether they are correct, so a strong-sounding argument is not evidence that the claim is true. It can also invent supporting statistics, which is why the prompt tells it not to — but you should still verify any figure before you cite it. Use this tool to map the landscape of an argument and surface trade-offs you had missed, not as an oracle that tells you which side is right. The judgement, with the cruxes in hand, stays yours.