CASE INTAKE · 001

Run Analysis

Enter your idea and the agent will search for evidence and try to kill it.

01Describe your idea in a sentence or two

02An AI agent searches Reddit, Product Hunt & the web for real evidence — not just its own guesses

03Get a scored KILL / PIVOT / BUILD report with an auditable formula, not just an opinion

Your Idea

LAB PROTOCOL · 002

Settings

Saved to the current browser session's server storage and only visible to you.

LLM Provider

Provider Model Base URL (optional)For Anthropic, use the root host like https://api.anthropic.com. For OpenAI-compatible gateways, use the gateway root URL, without /chat/completions.

Research Budget

Lower limits save tokens but may reduce research depth.

Max Tokens / request Max Agent Turns Max Searches Max Fetches Max Tool Calls

Default Run Parameters

Default Adventure Level

EVIDENCE ARCHIVE · 002

Reports

Only Pre-Mortem reports generated in this browser session are shown. Sessions are isolated per user.

← Select a report to view

ACCOUNT · 003

Account

Your credits balance and consumption history.

CREDIT ACQUISITION · 004

Buy Credits

Top up your balance to keep running the agent. 1 credit is consumed per run.

1 credit is consumed per run. It is only refunded if the run fails due to a system error — not if you cancel it yourself.

Credits balance

—

1 credit is consumed per run.

Purchases are processed securely by Creem. You will be redirected to a secure checkout page. Prices are in USD and excluding taxes.

By purchasing credits you agree to our Terms of Service and Privacy Policy. Prices exclude VAT and other applicable taxes.