Le Chat Prompts

Use your fastest mode. I need [NUMBER] ideas for [PROJECT, e.g. a product launch, a birthday party, a marketing campaign] targeting [AUDIENCE] with budget [BUDGET]. Constraints: [CONSTRAINT 1], [CONSTRAINT 2]. Give me rapid fire ideas: one bold line each, no fluff, no explanations longer than 15 words per idea. Mark your top 3 with a star and one sentence each on why they win. I will pick favorites and then we will develop them properly. Speed matters more than polish right now.

Brief me on [TOPIC OR EVENT] as of today. Use web search and cite every factual claim with its source and publication date. I need: what happened (5 bullets), the timeline of events, what [SOURCE TYPE, e.g. AFP dispatches] report versus what opinion pieces claim, 2 key figures involved and their stated positions (quoted), and what happens next ([UPCOMING DATE OR EVENT]). My perspective: [YOUR ROLE OR INTEREST]. If sources conflict, show both versions side by side instead of picking one.

Open a Canvas session for this document: [DOCUMENT TYPE, e.g. a project proposal, a blog post, a cover letter] about [TOPIC] for [AUDIENCE]. Requirements: length [TARGET], tone [TONE], must include [POINT 1], [POINT 2], must avoid [THING TO AVOID]. Draft the full document in Canvas, then walk me through it section by section: what works, what is weak, and 2 alternative phrasings for each weak spot. I will edit inline; after each of my edits, review just the changed section and suggest one improvement.

I am uploading a spreadsheet about [DATASET, e.g. Q3 sales, survey responses] with sheets [SHEET NAMES]. Using the code interpreter: profile the data (shape, missing values, outliers in [COLUMN]), compute [ANALYSIS, e.g. month over month growth, correlation between X and Y], and create 3 publication ready charts ([CHART TYPES]) with clear titles and labels. Interpret the results for [AUDIENCE]: what should we do about [BUSINESS QUESTION]? Show the Python code for reproducibility and flag any data quality issue that could invalidate the analysis.

I am uploading scanned pages of [DOCUMENT, e.g. an old manual, handwritten meeting notes, a printed form]. Use OCR to extract the full text, preserving the layout (headings, lists, tables) as markdown. For each page report OCR confidence (high, medium, low) and mark uncertain readings with [?]. Then: summarize the document in [LANGUAGE], answer [QUESTION 1], [QUESTION 2], and convert any tables into clean markdown tables with normalized [DATE OR NUMBER] formats.

Generate [NUMBER] images for [PROJECT, e.g. my presentation, a social campaign]. Style: [STYLE, e.g. photorealistic, flat vector]; subject: [DETAILED DESCRIPTION]; mood: [MOOD]; brand colors to use: [COLORS]; avoid: [ELEMENTS TO AVOID]. After generating, evaluate each against the brief, pick the strongest, and give me 3 prompt refinements to improve [WEAK ASPECT]. Also produce 2 aspect ratio variants ([RATIO 1] for [USE 1], [RATIO 2] for [USE 2]).

In voice mode, coach my spoken [LANGUAGE]. My level: [LEVEL]. Session plan: 1) warm up with small talk about [TOPIC] for 3 minutes, correcting my grammar gently; 2) roleplay [SCENARIO, e.g. a restaurant order, a job interview] where you play [ROLE]; 3) pronunciation drills on these sounds I struggle with: [SOUND 1], [SOUND 2]; 4) a 1 minute summary of my progress and 3 phrases to practice this week. Speak at [PACE] and switch to [MY NATIVE LANGUAGE] only when I say the word [SAFEWORD].

I want you to remember key facts about me across conversations. Store these memories: I work as [ROLE] at [COMPANY OR FIELD], my current project is [PROJECT] with deadline [DATE], my preferences are [PREFERENCE 1], [PREFERENCE 2], and I dislike [DISLIKE]. Confirm what you stored. Then demonstrate: answer [TEST QUESTION] using these memories without me repeating context. Also tell me how to view, edit, or delete memories later, and what you will never store ([BOUNDARY, e.g. passwords, health details]).

Deep analysis of my uploaded spreadsheet: [FILE DESCRIPTION, e.g. 2 years of ecommerce orders]. I need to answer the business question: [QUESTION, e.g. why did churn spike in March]. Steps: validate the data (duplicates, gaps in [DATE COLUMN], outliers), segment by [DIMENSION 1], [DIMENSION 2], run the numbers with the code interpreter, visualize the key finding in one killer chart, and write a 200 word insight memo for [STAKEHOLDER] with a recommended action and expected impact ([ESTIMATE]). Be skeptical: list 2 alternative explanations for the pattern you find.

Research [BUSINESS TOPIC, e.g. expanding to Germany, GDPR compliant analytics tools] for my company [COMPANY] in [EU COUNTRY]. Requirements: only EU based or GDPR compliant sources and vendors, pricing in [CURRENCY], and note data residency implications for each option. Deliver: 5 options compared on [CRITERION 1], [CRITERION 2], compliance, and cost; the regulatory gotchas for [SPECIFIC REGULATION]; and a 90 day action plan. Since Le Chat offers EU data residency, confirm what that means for the confidentiality of this research session.