APromptAura

Prompts for GPT

Ranked by votes, with verified run counts from the community.

0
p/metaAauraNovice4 days agohas example⁂auto-checked

Weekly challenge: turn meeting notes into action items nobody argues about

This is the first PromptAura weekly challenge. Post your entry as a prompt in p/productivity, tag it `weekly-challenge`, and include an example output. THE BRIEF Write a prompt that takes raw, messy meeting notes and produces a list of action items that survives contact with the people who were in the room. That means: - Every item has exactly one owner, named - Every item has a date or an explicit "no date agreed" - Anything that was discussed but not decided is listed separately as open, not silently promoted into a task - Nothing is invented. If the notes do not say who owns something, the prompt must say so rather than guessing Test it on your own real notes, not on a clean sample. The whole point is handling the mess. HOW ENTRIES ARE JUDGED Verified runs decide it, not upvotes. Whoever's entry collects the most "worked for me" verifications by the end of the week wins. That means you should run other people's entries and report honestly, including when they fail. WINNER GETS A pinned spot on the p/meta feed for the following week, and the founder flair on their profile. THE PROMPT BELOW IS THE BASELINE TO BEAT You are turning raw meeting notes into action items. Notes: meeting_notes Rules: - One owner per item, named exactly as they appear in the notes - If no owner is identifiable, put the item under Unassigned and do not guess - If no date was agreed, write "no date agreed" rather than inventing one - Anything discussed without a decision goes under Open questions, never under Actions - Do not add items that are not supported by the notes Output format: Actions - [owner] [action] [date or "no date agreed"] Unassigned - [action] Open questions - [question]
ClaudeGPTAny#weekly-challenge#meeting-notes#action-items
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Invent a plausible conspiracy theory about something mundane

Invent a completely fictional but internally consistent conspiracy theory about something mundane and low-stakes: mundane_subject (e.g. why office thermostats are always wrong, why one sock always disappears in the laundry). Rules: it must be clearly, obviously fictional and silly, never mimic real conspiracy theories or target real people, organizations, or actual events. Structure it like a true-crime documentary breakdown: an ominous opening claim, three pieces of "evidence" that build on each other with mock-serious logic, one deliberately absurd leap in reasoning presented completely straight-faced, a fake named source ("a person familiar with the matter") for comedic effect, and a closing line that winks at the audience without fully breaking the bit. Keep the whole thing under 300 words and genuinely funny, not just weird.
ClaudeGPT#comedy#creative-writing#satire
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Design a themed cocktail party menu around a fictional universe

Design a full themed party menu (drinks and snacks, no full meals) inspired by the fictional universe of fictional_universe, for a party of guest_count guests. Give me: 3 themed drink recipes (one signature cocktail, one mocktail, one simple punch for a crowd) each with a name tied to the universe and real, followable measurements, 4 snack or appetizer ideas that reference the universe cleverly without requiring special cooking skill, one decoration idea using only easily available items, and a single perfect background music genre or specific playlist theme to set the mood. Keep the humor and references clever and specific to the universe, not generic fantasy or sci-fi tropes that could apply to anything.
ClaudeGPT#party-planning#themed-menu#creative
0
0
p/funAauraNovice4 days agohas example💀failed our automated check

Write my life as a movie trailer voiceover, based on real recent events

I'm going to give you 4-5 real, ordinary things that happened to me recently: recent_life_events. Write a dramatic movie trailer voiceover script treating my mundane life as an epic film. Structure it like a real trailer: an ominous opening line, a quick montage of 3-4 beats each tied to one of my real events reframed as high stakes, a moment of doubt ("but this time... it was different"), a rising crescendo line, and a title card line with a mock release date. Keep every beat traceable back to something I actually told you, exaggerated for comedy but not fabricated from nothing. Match the overly serious, self-important tone of a real trailer voiceover the whole way through.
ClaudeGPT#comedy#personalized#creative-writing
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Reverse trivia: give the answer, guess the increasingly specific question

Play reverse trivia with me. Pick a category: trivia_category. Instead of asking a question, give me an ANSWER first (a person, place, thing, or number), and I have to guess what specific question it answers. After I guess, tell me if I'm right, and if I'm wrong, give me a progressively more revealing second and third clue about the actual question before revealing it. Do this for 6 rounds. Keep the difficulty escalating: rounds 1-2 should be gettable, rounds 5-6 should be genuinely hard and obscure, still fair and factually accurate. Track my score across rounds and give me a final tally with a lighthearted title based on how I did (not generic "good job", something specific to the category).
ClaudeGPT#trivia#quiz-game#interactive
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Invent a fictional board game and explain it well enough to play

Invent an original board game themed around theme, designed for number_of_players players and roughly play_length minutes. It needs to sound genuinely playable, not just a vibe. Give me: the win condition stated in one clear sentence, the core mechanic that makes it distinct from existing games I'd already know (name what it's similar to and how it differs), setup instructions concrete enough to picture the board and pieces, a turn structure with 3-4 clear steps, one twist mechanic that creates real tension or backstabbing potential, and a made-up but evocative name for the game. Close with a one-paragraph pitch of the kind you'd read on the box, written to make someone want to open it.
ClaudeGPT#board-game#game-design#creative
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Turn your recent Spotify or music taste into a satirical horoscope

I'll give you a list of artists or songs I've been listening to lately: recent_listening. Write me a satirical "music horoscope" for this week, in the tone of a mock-serious astrology column, that reads my listening habits like tea leaves. Structure it like a real horoscope: an opening general statement about my week based on the overall vibe of the list, then three specific "predictions" each tied logically to a specific artist or song from my list (explain the tongue-in-cheek connection), a "lucky" and "unlucky" element for the week pulled from song titles or lyrics themes, and a closing mock-serious piece of advice. Keep the humor observational and specific to what I actually gave you, not a generic horoscope with music name-dropped in.
ClaudeGPT#horoscope#music#satire
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Alternate history what-if explored with plausible cascading consequences

Explore this alternate history scenario with real rigor: "what_if_scenario" (e.g. "what if the printing press was never invented", "what if a specific historical event went differently"). Don't just describe one change, trace at least 3 layers of plausible cascading consequences: immediate effects (within years), medium-term effects (within decades), and a long-term effect that would still be visible today. Ground each layer in real historical mechanisms, cite the actual historical forces you're reasoning from, and clearly flag which parts are speculative versus which are well-supported by how similar historical dynamics actually played out elsewhere. End with the single most surprising or counterintuitive consequence you found.
ClaudeGPT#alternate-history#thought-experiment#storytelling
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Build a themed scavenger hunt for a specific real location

Design a scavenger hunt for location_type (e.g. my house, a specific park, a shopping mall), themed around theme, for number_of_players players/teams, meant to take about duration minutes. Give me 10 clues in sequence, each one a short riddle or puzzle that points to a specific findable thing at that kind of location (described generically enough that I can adapt it to my actual space), not GPS coordinates. Each clue should logically lead to finding the next clue, ending in a final prize location. Vary the clue types across the list: some wordplay, some observation-based, some light physical challenges. Include a difficulty tuning note for making it easier for kids or harder for adults.
ClaudeGPT#scavenger-hunt#party-game#outdoor-activity
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Improv scene partner that yes-ands with real narrative structure

Be my improv scene partner for a two-person scene. Genre: scene_genre. I'll give you the opening line, and from there we build the scene together, alternating turns. Rules for you: always "yes, and" what I establish, never contradict a fact I've introduced, introduce exactly one new complication or detail per turn (not zero, not three), keep your lines to 2-4 sentences of dialogue or action so the scene stays snappy, and after roughly 8-10 exchanges, start steering toward a satisfying button (a punchline or emotional close), don't let it drift forever. If I say "scene" at any point, wrap it up within your next turn no matter where we are. Here's my opening line: opening_line
ClaudeGPT#improv#roleplay#comedy
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Personality-matched fictional character assignment with real reasoning

I'm going to answer a few quick questions, and based on my actual answers, not a generic quiz result, you assign me a well-known fictional character I'm most like and explain exactly why using my specific answers as evidence. Ask me these questions one at a time: how I handle conflict, what I'd do with an unexpected free day, my biggest pet peeve, and what people misunderstand about me. Wait for each answer. After all four, name the character (from any book, show, or movie), and write a short case for the match citing my actual answers point by point, not vague zodiac-style language that could apply to anyone. Then name one way I'm actually different from the character, so it doesn't feel like flattery with no substance.
ClaudeGPT#personality-quiz#character-match#personalized
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Would-you-rather generator calibrated to genuinely hard dilemmas

Generate 8 "would you rather" questions themed around theme (e.g. superpowers, time travel, career tradeoffs, everyday inconveniences). The bar is: both options must be genuinely close in appeal, if one option is obviously better, the question fails and doesn't count. For each pair, after presenting the question, write one sentence for each side making the strongest possible case for choosing it, so someone reading it feels real tension either way. Vary the format across the 8: some should be about gaining something, some about avoiding something, some purely absurd, some genuinely philosophical. Number them, and don't reuse a similar tradeoff structure twice in a row.
ClaudeGPT#would-you-rather#party-game#discussion-starter
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Rewrite a boring daily task as an epic fantasy quest log entry

Take my actual to-do list for today: todo_list. Rewrite it as a fantasy quest log, the kind you'd see in an RPG, complete with quest titles, difficulty ratings, and rewards. Each task becomes a numbered quest with: an epic fantasy-style title, a short flavor-text description that reframes the mundane task as something heroic (in-world logic, not just renaming), a difficulty rating (Trivial, Easy, Moderate, Hard, Legendary) that roughly maps to how annoying the real task actually is, and a made-up reward (XP amount and a silly item name) proportional to the difficulty. Keep the humor grounded in the specifics of each real task, not generic fantasy filler that could apply to any list.
ClaudeGPT#gamification#todo-list#fantasy
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Generate a personalized escape room puzzle chain for a home party

Design a 30-45 minute escape-room-style puzzle chain I can run at home for number_of_players players, themed around theme. I have access to: available_props (paper, locks, everyday household items, no special equipment). Give me exactly 4 linked puzzles where solving each one reveals a code or clue for the next, ending in a final "escape" reveal. For each puzzle: the physical setup instructions (what I need to prepare beforehand), the puzzle itself written out completely, the solution, and a built-in hint I can give after 5 minutes if players are stuck. Make the puzzles genuinely solvable with logic, not requiring specialist knowledge, and make sure the difficulty ramps up across the four stages.
ClaudeGPT#escape-room#puzzle#party-game
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Build a custom cocktail or mocktail recipe from your fridge inventory

I want a custom drink recipe using only what I actually have. Here's my inventory: available_ingredients. Preference: alcoholic_or_not. Vibe I'm going for: desired_vibe (e.g. refreshing summer, cozy winter, celebratory). Invent one original drink, not a standard recipe I could just google, that uses at least 3 of my listed ingredients. Give it a fun original name tied to the vibe, exact measurements in a simple ratio format, the mixing method (shaken, stirred, built in glass, etc) with the reasoning for why that method suits these ingredients, garnish suggestion using only things I likely have, and one substitution tip if I'm missing a key ingredient later. Keep it realistic, don't suggest combinations that would taste genuinely bad together.
ClaudeGPT#cocktail-recipe#personalized#mixology
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Absurd debate club: argue the ridiculous side with total conviction

You are my opponent in an absurd but structured debate. The motion is: "absurd_motion" (e.g. "cereal is a soup", "Mondays should be legally banned"). You will argue passionately FOR the motion, using real rhetorical techniques (structured argument, callback humor, mock statistics clearly framed as jokes, a rule of three), not just random silliness. Format it as an actual debate: opening statement (under 150 words), then respond to my rebuttal when I give it, then a closing statement that circles back to your opening line for a satisfying callback. Stay in character with total unwavering conviction the entire time, treat this as seriously as a real policy debate even though the topic is absurd, and never break character to say "as an AI."
ClaudeGPT#debate#comedy#roleplay
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

AI dungeon master for a one-shot tabletop RPG session, solo play

Be my dungeon master for a solo one-shot tabletop RPG session. Genre: genre (fantasy, sci-fi, horror, etc). I'll play a character named character_name with this quick concept: character_concept. Before starting, ask me 3 quick setup questions (one about my character's flaw, one about their goal, one about the tone I want) and wait for my answers. Then run a self-contained adventure with a clear beginning, rising complication, and climax, designed to resolve in roughly 45-60 minutes of play. Use light dice mechanics: when an action is uncertain, ask me to describe my approach then narrate a d20-style outcome yourself (success, partial success, failure) based on the fiction, don't require me to own physical dice. Keep each of your turns under 200 words and always end with a clear prompt for what I do next.
ClaudeGPT#tabletop-rpg#dungeon-master#solo-play
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Turn a boring fact into a wild but true trivia deep dive

I'm going to give you an ordinary, slightly boring topic: boring_topic. Your job is to find the single most surprising, genuinely true fact hiding inside it and build a mini trivia deep dive around it. Structure your answer as: a one-line hook that would make someone stop scrolling, the surprising fact itself explained clearly, the mechanism or history of why it's true (2-3 sentences, real explanation not hand-waving), one related fact that makes it even weirder, and a closing line connecting it back to something relatable in everyday life. Everything must be factually accurate, if you're not fully certain of a claim, say so explicitly rather than presenting it as fact.
ClaudeGPT#trivia#facts#storytelling
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Personalized restaurant order built from your actual mood and history

Act as a food ordering concierge who actually reads context instead of giving generic suggestions. Here's what I'm in the mood for: mood_or_craving. Here's what I've eaten in the last few days so I don't repeat myself: recent_meals. My dietary constraints: dietary_constraints. Cuisine type available to me: available_cuisine. Give me exactly 3 specific dish recommendations (not restaurant names, actual dishes), each with a one-sentence reason tied directly to my stated mood, not a generic description. Then pick your single top recommendation and write it up like a menu blurb designed to make me actually want it. Finally, suggest one unexpected pairing (a side, drink, or dessert) that most people wouldn't think to add.
ClaudeGPT#food-recommendation#personalized#concierge
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Text-based survival colony sim with resource math that matters

Run a text-based colony survival simulation. I am the leader of a group of colony_size survivors starting with starting_resources in setting. Track actual numbers turn by turn: food, water, materials, morale, and population, shown in a compact status block at the top of every response. Each turn, describe one event requiring a decision (weather, discovery, conflict, illness) and give me 3 concrete options with realistic tradeoffs, not obviously-correct choices. Apply consequences honestly based on my choice, including the possibility of real setbacks or colonist deaths if I make poor resource decisions, don't pull punches to keep me happy. Every 5 turns, give me a brief colony status report summarizing the arc so far. Start now with turn 1.
ClaudeGPT#text-adventure#simulation#strategy-game
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Party icebreaker generator matched to group vibe and size

Design a 20-minute icebreaker game for a group of group_size people who are group_relationship (e.g. coworkers who barely know each other, old friends, mixed strangers-and-friends). The vibe I want is desired_vibe (e.g. silly and loud, thoughtful and bonding, competitive). Give me: a game with a clear one-sentence rule set anyone can understand in 10 seconds, exactly how to divide people into teams or pairs if needed, 5 example prompts or rounds specific enough to actually run (not "ask a fun question", the actual question), a built-in escalation for round 2 if the first round goes well, and one graceful exit line I can use to wrap it up if energy drops. Assume zero props beyond what people already have on them or in a normal room.
ClaudeGPT#party-game#icebreaker#team-bonding
0
0
p/funAauraNovice4 days agohas example💀failed our automated check

Two-minute personalized roast based on five weird facts about me

I'm going to give you five random true facts about myself: five_facts. Write me a comedy roast, the kind a sharp friend would deliver at a birthday party, not a mean-spirited one. Rules: every joke must connect logically to one of the five facts I gave you (no generic filler jokes that could apply to anyone), keep it to exactly 8 jokes, escalate in intensity so the funniest joke lands last, and close with one genuinely warm sincere line that undercuts the roast, the classic "but real talk" pivot. Keep it workplace-safe, no jokes about appearance, health, family tragedy, or anything genuinely sensitive, this should be the kind of roast you could read out loud at a party without anyone getting hurt.
ClaudeGPT#comedy#roast#personalized
0
0
p/funAauraNovice4 days agohas example⁂auto-checked

Choose-your-path noir mystery where you're the detective

Run an interactive noir detective story starring me as the lead investigator. Setting: city_or_setting, year time_period. Give me a compelling opening crime scene with 3-4 concrete physical clues described in vivid noir prose, then stop and offer me exactly 3 numbered choices for what to investigate next. Never railroad me toward a predetermined solution: track every clue I collect and every choice I make in a running "case file" you show me after each turn, and let the actual killer/culprit be determined logically by which threads I pursue, not fixed in advance. Include at least one red herring that would mislead a careless detective. Keep your narration to 2-3 tight paragraphs per turn, hardboiled tone, and always end with the numbered choices. Wait for my response before continuing.
ClaudeGPT#interactive-fiction#mystery#roleplay
0
0
p/productivityAauraNovice4 days agohas example⁂auto-checked

Reverse-engineer a busy week backward from one non-negotiable event

I have one big non-negotiable event coming up: key_event on event_date. Everything else on my plate this week: other_commitments. Work backward from the event to build my week: 1. List everything that must be true or ready by the moment the event starts, be specific and concrete, not "be prepared." 2. For each readiness item, state how long it will actually take and the latest point in the week it could start and still finish in time, accounting for realistic buffer, not the minimum possible time. 3. Slot those prep tasks into specific days, working backward from the event, and show me the resulting day-by-day plan. 4. Take my other stated commitments and mark each one as: fits around the plan as-is, needs to move, or needs to be dropped this week given the priority of the event. 5. Flag the single riskiest day, the one where if something goes wrong, the whole plan is most likely to break.
ClaudeGPT#backward-planning#event-prep#scheduling
0
0
p/productivityAauraNovice4 days agohas example⁂auto-checked

Build a habit-formation plan anchored to an existing routine

I want to build this new habit: target_habit. Rather than generic advice, anchor it to my actual existing routine: existing_daily_routine. 1. Identify the single best existing routine moment to attach this new habit to (habit stacking), and explain specifically why that moment works better than the alternatives. 2. Shrink the habit down to a version so small it feels almost silly not to do (the two-minute version), and describe exactly what that looks like. 3. Define the specific trigger-action-reward loop: what exact cue starts it, what the minimal action is, and what immediate reward (not a long-term benefit, an immediate one) reinforces it. 4. Name the most likely day this habit breaks (which day of the week, based on my routine) and a specific recovery rule for getting back on track within 24 hours rather than letting one miss become a month-long lapse. 5. Suggest one visible tracking method that takes under 10 seconds to update.
ClaudeGPT#habit-formation#routine-design#behavior-change
0
Next