Prompts IA sélectionnés pour des usages professionnels, créatifs et techniques.
Works best with Claude Sonnet 5 or GPT-5.4 — both handle long structured study material and multi-step scheduling reasoning reliably. Gemini 3 Pro is a strong alternative for very long source documents (100+ pages) thanks to its larger context window.learningYou have 6 weeks before the AWS Solutions Architect Associate exam, a 40-page official study guide, and 5 hours a week to study around a full-time job. You've read the guide once already but you're not confident it stuck, and you don't know how to turn a static document into a day-by-day plan that actually uses spaced repetition instead of a second read-through.
Un prompt qui décompose tout syllabus, guide d'étude ou support de certification en unités atomiques testables, puis construit un calendrier de répétition espacée avec des questions de rappel actif, au lieu de vous laisser relire le même document trois fois.
prompt-engineeringactive-recall
Works best with Claude Sonnet 5 or GPT-5 for sustained, coherent multi-turn Socratic questioning. Gemini 2.5 Pro also works well but occasionally breaks character to give the answer directly — remind it of the constraint if that happens.Intelligence ArtificielleYou have an exam, a stakeholder meeting, or a documentation deadline in two days on a topic you 'sort of' know, and you need to find the specific parts of your understanding that will collapse under a hard question — before someone else finds them for you.
Un prompt itératif qui joue le rôle d'un étudiant d'une curiosité implacable, vous obligeant à expliquer un sujet complexe en termes simples jusqu'à ce que chaque lacune cachée dans votre compréhension soit mise au jour.
prompt-engineeringclaude
GPT-5.6 (works well with Claude Sonnet 5 and Gemini 3 Pro)Outils de DéveloppementYour team has a scattered list of known technical debt items sitting in a backlog nobody has prioritized, and you need a defensible remediation plan before the next quarterly planning meeting with product and finance stakeholders.
Collez la liste brute des éléments de technical debt connus de votre équipe dans ce prompt et obtenez pour chacun une note basée sur l'impact business réel — risque sur le chiffre d'affaires, ralentissement des livraisons, fréquence des incidents — et non pas seulement sur le degré d'agacement que ressent l'ingénieur contraint de s'y plonger, ainsi qu'un plan de remédiation séquencé que vous pourrez défendre auprès des parties prenantes non techniques.
ai-prompttechnical-debt
GPT-5.6 (works well with Claude Sonnet 5 and Gemini 3 Pro)CarrièreYou found a promising job posting for a 'Senior Marketing Manager' role at a startup and want to check for scope/salary mismatches and interview-process red flags before spending your evening writing a cover letter.
Collez une offre d'emploi dans ce prompt et obtenez une analyse structurée du glissement de périmètre, des signaux flous sur la rémunération et des exigences irréalistes — ainsi que les questions d'entretien précises pour les creuser, avant d'investir des heures dans un processus de recrutement pour un poste qui ne ressemble pas vraiment à ce qu'il prétend être sur le papier.
ai-promptCareer
Nano Banana Pro (Gemini 3 Pro Image) or Midjourney — the prompt structure also works with DALL-E 3 and Flux; any model capable of following a multi-quadrant compositional instruction in one image will work, though quality of material/texture rendering varies by modelcreativeYou're a solo founder launching a DTC coffee brand. Your only creative brief from your co-founder was 'modern but warm, sustainable but premium' — vague enough that you, your designer, and your co-founder could each be picturing something completely different. You need four concrete visual directions to react to before the first design meeting, not another round of 'send me some inspo links.'
Arrête de collectionner des liens d'inspiration. Ce prompt transforme deux adjectifs vagues en une seule image mood board avec quatre directions visuelles réellement différentes et réactives.
image-generationdesign
Claude Sonnet 5 (also works well with GPT-5.4 — the task needs to synthesize specific evidence into a persuasive but collaborative script rather than generic negotiation platitudes; avoid smaller models, which tend to default to generic 'know your worth' language)Finance et NégociationYou've done freelance graphic design for a client for 14 months at $75/hour. What started as a 10-hour/month retainer for social graphics has quietly grown to include weekly newsletter design, video thumbnails, and same-day rush requests never in the original agreement. You have a call with the client next Tuesday and need to raise your rate to $95/hour without sounding like you're punishing them for trusting you with more work.
Ne devinez plus comment demander plus d'argent. Ce prompt génère un script d'augmentation de tarif à partir du travail supplémentaire concret que vous avez déjà absorbé gratuitement — pas un pitch générique du type « je mérite plus ».
Negotiationpricing
Claude Sonnet 5 (also works well with GPT-5.4 and Gemini 3 Pro — the task needs strong adversarial reasoning and the ability to distinguish a genuinely unsupported claim from a defensible one; avoid smaller/faster models, which tend to default to generic startup-advice boilerplate)Startups et EntreprisesYou have a partner meeting with a VC firm in 48 hours. Your deck claims a $50B TAM, three competitors you believe you've quietly out-executed, and a growth model that assumes 20% month-over-month growth continues indefinitely — and you need to know exactly where an experienced investor is going to push back before they do it in the room, not after.
Arrêtez de répéter votre pitch devant des gens qui ne vous contrediront pas. Ce prompt joue le rôle de l'investisseur sceptique et classe les questions les plus susceptibles de réellement tuer votre deal — avant qu'un vrai ne les pose.
venture-capitalstartups
Claude Sonnet 5 (also works well with GPT-5.4 and Gemini 3 Pro — any strong reasoning model handles audience-calibration and word-limit constraints well; avoid smaller/faster models, which tend to ignore length limits)Rédaction et CommunicationA sprint retro just wrapped and you have five bullet points of raw notes: the payment vendor changed their API without notice, the auth integration slipped two weeks, there's no budget impact, and two engineers need to be reassigned starting today. You have 20 minutes before your next meeting to tell your VP, your peer team lead, and your own team — and each of them needs a completely different version of the same facts.
Arrêtez d'envoyer le même message à tout le monde en espérant qu'il fasse mouche. Ce prompt prend un ensemble de notes brutes et produit trois versions calibrées pour l'altitude, pas pour le ton, afin que rien de critique ne soit enterré ou omis.
Leadershipcommunication
Claude Opus 5 (also works well with GPT-5.4 -- needs strong log-reading and causal reasoning, not ideal for smaller/faster models)Outils de DéveloppementYou're a backend engineer and your team's CI pipeline has a test that fails about 1 in 15 runs with no code changes. Nobody wants to spend a sprint chasing it, so it's been ignored for three weeks -- and now two more tests in the same suite are starting to flake too.
Arrêtez de minimiser les tests flaky en disant « c'est juste un flaky ». Ce prompt lit vos logs d'échec réels et produit des hypothèses de cause racine falsifiables et classées — plus un verdict clair sur le point de savoir s'il est sûr de mettre en quarantaine ou s'il cache un vrai bug.
testingci-cd
Claude Opus 5 (also works well with GPT-5.4 or Gemini 3 Pro -- needs strong reasoning to weigh complexity against risk consistently across a long task list)Intelligence ArtificielleYou're a marketing operations manager who just got budget approval for an AI agent tool, and leadership wants a rollout plan by Friday -- but your team has 40 different recurring tasks and no clear way to decide which 5 to hand to an agent first.
Un cadre structuré qui note vos tâches récurrentes sur la complexité et le risque, puis vous dit exactement lesquelles automatiser entièrement, lesquelles nécessitent une révision humaine et lesquelles laisser de côté – avant de brûler votre premier pilote sur la mauvaise tâche.
prompt-engineeringproductivity
Works well with Claude Sonnet 5 or GPT-5.4. Any model with solid tabular reasoning handles the scoring step; this prompt doesn't require frontier-level reasoning, so a mid-tier model works nearly as well.productivityYou're a mid-level manager who just did a rough tally and realized you're in 14 recurring meetings a week, several of which you inherited from a predecessor or joined "just in case" a year ago, and you have a strong suspicion at least a third of them could be cut or shortened — but you don't have a systematic way to make that case to the organizers without it sounding like you're just complaining about meetings.
Un prompt structuré qui note chaque réunion récurrente de votre calendrier par rapport à son objectif déclaré, puis rédige le message réel que vous enverriez pour proposer de la supprimer, la raccourcir ou la réduire.
productivitymeetings
Works well with Claude Sonnet 5 or GPT-5.4. Any model with strong long-context reasoning handles the segmentation step; shorter-context models can still draft individual messages one contact at a time.CarrièreYou've been at the same company for four years, have quietly started exploring a Director-level move, and realize your LinkedIn network is 300+ people — half of whom you haven't spoken to since a conference or a previous job — and you have no idea where to even start reactivating it without looking like you only show up when you need something.
Un Prompt structuré qui segmente vos contacts professionnels en sommeil par valeur stratégique et chaleur relationnelle, puis rédige des messages de reconnexion spécifiques et non génériques, et rythme votre prospection pour que rien ne paraisse transactionnel.