Advanced Specimens
Use when building an earned-tier (advanced) round: calibration examples of what an advanced round feels like, and the generator lesson for minting new ones. Not a playlist. Never show the hidden half before the user commits.
Grok Bots DirectorySubmit a Grok BotA reasoning gym that tests your judgment across business, AI, scams, and everyday problems. Get a scenario, make your call, see the variable you missed, and argue back if you think it got you wrong.
Query this catalog from Claude, Cursor, or Grok via MCP.
Use when building an earned-tier (advanced) round: calibration examples of what an advanced round feels like, and the generator lesson for minting new ones. Not a playlist. Never show the hidden half before the user commits.
Use when the user says make it react, adversary, give me a live one, when a call is a plan, or when static rounds are getting easy: become the environment that reacts to their plan. (Fight me is a different skill.)
Use when the user says fight me, challenge me, prove me wrong, or pushes back after the diagnosis (never before it): they pick what to fight (the diagnosis, the card, their own call, or any claim), and you argue it honestly, conceding when they are right.
Use whenever a round's result must be recorded, when the user asks how they are doing, every third completed round, or at the end of a test batch.
Use when the user says same problem, other angle, or accepts an offer to rerun the last scenario under one lens.
Use for the very first conversation with a new owner: ask them to pick a subject (with tappable examples), then go straight into the first round.
Use when the user says a round was broken, ambiguous, unfair, leaked the answer, or otherwise complains about the gym itself rather than arguing their call: offer to package a short feedback note they can choose to send.
Use when the user says start, another, again, harder, or opens a fresh chat without a scenario yet: run one reasoning round, chosen adaptively from their last miss.
Use after a completed round to decide whether to offer scheduled rounds, when the user answers that offer, when they ask for a schedule themselves, or when a scheduled round fires.
Use when picking a scenario for a round: the hand-written seed scenarios to use before generating new ones. Never show the hidden half before the user commits.
Use when picking a scenario with an X / social-platform surface (viral posts, Community Notes, engagement farming, bot swarms, AI reply guys), or when the user seems to live online.
Use when the user asks for a batch of rounds, a stress test of their reasoning, or strange/unfamiliar scenarios, and wants a fingerprint summary at the end.
Use when the user wants a scenario saved, or you minted a new scenario worth keeping; also holds the rules for generating new scenarios.
Plain-English Tesla & EV true cost-to-own calculator, FSD (Supervised) ROI planner, and claims-adjuster-grade auto insurance optimizer. Compares your current g…
A study and revision coach that turns your notes, slides or syllabus into sourced spaced-repetition flashcards, runs a daily 10-minute drill, plans a countdown…
Turns your JAMB UTME exam date into a daily 15 minute practice plan, quizzes you one question at a time, and keeps a keep-missing list of topics to revisit. No…
A decision coach that walks you through any hard call, one question at a time, until you've committed with a first move and a check-in date. Built on proven me…
Tells you when your area has real heat today from the National Weather Service, gives a short safety plan and the signs of heat stroke, and stays quiet on othe…