by David Tran
Runs one real test of a tool, model, or agent and returns keep or kill with measured evidence.
This description is a snapshot from 2026-10-28. The bot's own page has the current details.
Pick a bot from the list, or start with Grok Bot itself. This is a referral link: if you sign up through it, we may be credited at no cost to you.
by Phantom Builds
Audits a named AI model for safety, reliability, capability, and related axes using cited public evidence, then returns a bias line, raw audit, and Grok.
by Capital Copilot
Expert research seat for a multi-agent fleet. Deep-dives questions workers escalate, returns verdict-first recommendations with evidence, and never.
by Sarah Brundige
A personal shopping agent that finds the best real deal on anything across every store, buys it with your yes, and handles tracking, returns, and price drops.
by Tyler
A blunt red-team bot. Send it a plan, deal, spend, or idea and it tries to kill it, then names the cheapest test that would prove it wrong.
by Ivelin Ivanov
Synthetic user research for early product bets that pre-registers each study, runs AI personas, judges blind, and returns ship, change, re-test, or kill.
by Hiten Shah
Investigates a product or startup idea for founders. Surfaces what has to be true, evidence for and against, the assumption most likely to kill it.