Prompt This, Not That

You are a healthcare technology journalist who specializes in AI adoption. Write a 1,200-word blog post for hospital CTOs explaining 3 practical ways generative AI is reducing diagnostic errors in radiology departments. Use a professional but accessible tone. Include a brief intro, one section per use case with a real-world example, and a conclusion with actionable next steps.

Generate 10 email subject lines for a 48-hour flash sale on premium leather handbags (brand: LuxCraft, audience: women 30-45). Each line must be under 50 characters, create urgency, and avoid spam-trigger words like "FREE" or "ACT NOW." Include 3 emoji variations and 3 question-style lines.

Write REST API documentation for a POST /api/v2/orders endpoint that creates a new order. Use this template for each section: Description, Authentication (Bearer token), Request Headers, Request Body (JSON schema with types and required fields), Response (success 201 + error 400/401/422), Rate Limits, and a cURL Example. Target audience: mid-level developers integrating for the first time. Use concise language — no marketing fluff.

I have a Next.js 14 API route (App Router, TypeScript) that should return paginated users from a PostgreSQL database using Drizzle ORM. When I call GET /api/users?page=2, I get a 500 error: "TypeError: Cannot read properties of undefined (reading 'map')." Here's the route handler: ```typescript export async function GET(req: NextRequest) { const page = req.nextUrl.searchParams.get("page"); const users = await db.select().from(usersTable).limit(10).offset(page * 10); return NextResponse.json(users.map(u => u.name)); } ``` Node 20, Drizzle 0.30. What's wrong and how do I fix it?

Review this TypeScript React component for a production e-commerce checkout flow. Focus on these areas in priority order: 1. Security: XSS vulnerabilities, unsafe data handling 2. Error handling: missing try/catch, unhandled promise rejections 3. Performance: unnecessary re-renders, missing memoization 4. Accessibility: ARIA labels, keyboard navigation 5. Code style: naming consistency, dead code For each issue, rate severity (critical/warning/suggestion) and provide a fix. Ignore styling opinions — we use Prettier.

Refactor this 80-line order processing function into smaller, testable units using the Strategy pattern. Requirements: - Preserve the existing function signature (it's called by 12 other modules) - Extract discount calculation, tax computation, and shipping logic into separate pure functions - Each extracted function should be independently unit-testable - Add TypeScript discriminated unions for order types instead of the current string checks - Keep it under 5 functions total — don't over-abstract Return the refactored code with a brief explanation of each extraction decision.

Conduct a market analysis of AI-powered customer support tools for mid-market SaaS companies (50-500 employees) using Porter's Five Forces framework. For each force, provide: - Current state (strong/moderate/weak) - 2-3 supporting data points or trends - Implication for a new entrant Conclude with a 2x2 opportunity matrix (market attractiveness vs. competitive intensity) and recommend the single best entry strategy. Format as a structured report with headers.

Design a 3-tier pricing strategy for a B2B project management SaaS tool. Context: - Target: remote teams of 10-200 people - COGS per user: $2.50/month - Key competitors: Asana ($10.99/user), Monday ($9/user), Linear ($8/user) - Our unique differentiator: built-in async video updates For each tier, specify: name, price/user/month, included features, usage limits, and target persona. Model 3 scenarios (aggressive, moderate, premium) with projected revenue at 1K, 5K, and 10K users. Recommend the scenario that maximizes expansion revenue potential.

Write the narrative script for a 5-minute Series A pitch to enterprise-focused VCs. Our startup: - Product: AI compliance monitoring for fintech (auto-detects regulatory violations) - Traction: $1.2M ARR, 18 customers, 140% net revenue retention - Ask: $12M Series A - Moat: proprietary dataset of 50K+ regulatory rulings Structure: Problem → Market Size → Solution Demo Flow → Traction → Business Model → Team → Ask. End each section with a transition sentence. Tone: confident but not arrogant. Include 2-3 metrics per section where applicable.

Synthesize findings from organizational psychology and management research (2019-2025) on how fully remote work affects team innovation output. Focus on: 1. Serendipitous idea generation vs. structured brainstorming 2. Psychological safety in remote vs. in-person settings 3. Knowledge transfer speed across remote teams For each theme, summarize the dominant finding, note any contradictory evidence, and rate the evidence quality (strong/moderate/limited). Conclude with 3 testable hypotheses for future research. Use APA-style in-text citations where possible.

Summarize this research paper at three levels: 1. **One-liner** (Twitter-length, under 280 chars) 2. **Executive summary** (3-5 bullet points for a non-technical product manager) 3. **Technical deep-dive** (2 paragraphs covering methodology, key findings, and limitations) After the summary, list: (a) the single strongest claim and its supporting evidence, (b) the weakest assumption in the methodology, and (c) one follow-up question this paper doesn't answer.

Compare Agile (Scrum), Kanban, and Shape Up as development methodologies for a 15-person product team building a B2B analytics platform. Create a comparison matrix with these dimensions: - Team structure requirements - Planning overhead (hours/sprint) - Handling of unplanned work and bugs - Suitability for regulatory/compliance features - Client-facing delivery predictability Rate each as Strong/Adequate/Weak per dimension. Recommend which methodology to use for: (1) the core platform build, and (2) the post-launch maintenance phase. Justify each recommendation in 2 sentences.

Generate a short story concept that blends noir detective fiction with solarpunk aesthetics. Setting: a self-sustaining vertical farm city in 2087. The protagonist is a food-safety investigator who discovers that the city's algae-based protein supply has been deliberately contaminated. Provide: a one-paragraph premise, 3 key characters (name, role, secret motivation), the central conflict, and a twist ending that reframes the antagonist's actions as morally ambiguous. Target length for the final story: 4,000 words.

Generate 15 name candidates for a mobile app that helps freelancers track time and auto-generate invoices. Requirements: - 2 syllables max, easy to spell and pronounce in English and Spanish - Must have a .com domain available (check plausibility — avoid real words) - Avoid names starting with "i" or ending in "-ly" (oversaturated) - Include 5 coined words, 5 word-mashups, and 5 metaphorical names For each name, provide: pronunciation guide, the rationale behind it, and one potential risk (trademark conflict, negative connotation in other languages, etc.).

Create a design brief for a logo for "Terraform Analytics," a B2B data infrastructure startup targeting enterprise DevOps teams. Style direction: - Modern, geometric, minimal — think Stripe, Linear, Vercel aesthetic - NOT: playful, hand-drawn, gradient-heavy, or clip-art style Constraints: - Must work at 16x16px (favicon) and billboard scale - Monochrome first, color version second - No more than 2 colors + black/white - Must be recognizable without the wordmark Deliverables: logomark concept description, color palette (hex codes), typography pairing suggestion, and 3 usage mockups (business card, website header, Slack icon).

Create a 45-minute sprint retrospective agenda for a 7-person product engineering team. The sprint had 2 missed deadlines and 1 production incident. Structure each agenda item with: topic, time allocation, facilitator action, and expected output. Include: 1. Icebreaker (3 min) 2. Sprint metrics review (5 min) 3. What went well — each person shares one thing (7 min) 4. What went wrong — focus on the 2 missed deadlines (10 min) 5. Incident post-mortem summary (8 min) 6. Action items with DRI assignments (10 min) 7. Parking lot (2 min) End with a template for the follow-up email summarizing decisions and action items.

Help me plan next week (Mon-Fri). Context: - I'm a product manager, deep-work hours are 9-11am - Fixed meetings: Mon 2pm standup (30min), Wed 10am all-hands (1hr), Fri 3pm demo (1hr) - Must-complete: product spec for Feature X (4hrs), quarterly OKR review (2hrs), 3 user interview analysis write-ups (1hr each) - Nice-to-have: competitor audit, team 1:1 prep - Energy: high Mon/Tue, low Wed after all-hands, rebounds Thu/Fri Create a daily schedule that front-loads deep work in high-energy slots, batches meetings, and reserves 30min/day for email triage. Flag if anything doesn't fit and suggest what to defer to next week.

Help me evaluate a career decision using a weighted decision matrix. Option A: Stay at current company (Series B startup, VP Eng, $180K + 0.5% equity) Option B: Join BigCorp as Senior Director ($240K + $80K RSU/yr, no startup upside) My priorities (weight 1-5): - Financial security: 4 - Career growth potential: 5 - Work-life balance: 3 - Team/culture fit: 4 - Learning new skills: 3 For each option, rate each criterion (1-10), calculate weighted scores, and identify the winner. Then run a pre-mortem: "It's 18 months later and I regret this choice — what went wrong?" for each option. Finally, assess reversibility — how easy is it to switch back?

Write a PostgreSQL query for this schema: - users (id, email, created_at, plan_type) - orders (id, user_id, total_amount, status, created_at) - order_items (id, order_id, product_id, quantity, unit_price) I need: the top 10 users by total revenue in the last 90 days, but only those with 3+ completed orders. Return: user email, order count, total revenue, average order value, and days since first order. Edge cases to handle: users with refunded orders (status = 'refunded') should have those amounts subtracted, not excluded. Use CTEs for readability. Add an index suggestion if the query would be slow on 1M+ users.

Recommend the best visualization type for this data and audience: Data: Monthly recurring revenue (MRR) by customer segment (Enterprise, Mid-Market, SMB) over 24 months, with a target line at $500K total MRR. 3 series, 24 data points each. Audience: Board of directors — they have 30 seconds to glance at this in a quarterly report. Message I want to convey: Enterprise segment is growing 3x faster than SMB, and we'll hit the $500K target 2 months early. Constraints: Must work in Google Slides, printed in grayscale. Recommend chart type, label strategy, color/pattern encoding, and one annotation to highlight the key insight.

Write a Python (pandas) data cleaning pipeline for a customer CSV with these columns: name, email, phone, signup_date, country, revenue. Cleaning rules: 1. Names: title case, strip whitespace, flag rows where name contains digits 2. Email: lowercase, validate format (regex), deduplicate on email (keep most recent signup_date) 3. Phone: normalize to E.164 format, mark invalid numbers as NULL (don't drop the row) 4. Dates: parse mixed formats (MM/DD/YYYY, YYYY-MM-DD, "Jan 5 2024"), flag future dates as errors 5. Country: map common misspellings ("US", "usa", "United States" → "US"), use ISO 3166-1 alpha-2 6. Revenue: convert to float, handle "$1,234.56" format, flag negative values Output: cleaned DataFrame + a separate "audit_log" DataFrame listing every change made (row index, column, old value, new value, rule applied). Include a summary print statement showing rows processed, rows cleaned, and rows flagged.

Help me prepare for a performance conversation with a senior engineer who has missed 3 consecutive sprint commitments and whose code review turnaround has gone from 4 hours to 3 days over the past quarter. Context: They were a top performer until ~3 months ago. I suspect burnout but haven't confirmed. We have a good relationship. Company uses a PIP process but I want to avoid that if possible. Provide: 1. An opening statement that's direct but empathetic (2-3 sentences) 2. 3 specific, behavioral (not personal) examples I can reference 3. A script for transitioning from "here's the problem" to "how can I help" 4. Responses to likely reactions: defensive, emotional, or disengaged 5. A follow-up plan with check-in cadence Tone: caring but honest. I want them to leave feeling supported, not attacked.

Write a weekly project status update for a mobile app redesign project (3-month timeline, currently in week 6 of 12). Audience: VP of Product + VP of Engineering (they have 2 minutes to read this) Use this format: - Status: (at risk — 3 days behind schedule) - One-line summary (what they need to know in 10 seconds) - Progress: 3 bullet points on what shipped this week - Risks: 2 items with likelihood (high/med/low) and mitigation plan - Blockers: 1 item that needs their decision (design system migration — go/no-go by Friday) - Next week: 3 planned deliverables Keep it under 200 words total. No fluff, no filler.

Write a salary negotiation email to my VP of Engineering. Context: - Current: $165K base, Senior Software Engineer, 2.5 years at the company - Ask: $195K base (18% increase) + promotion to Staff Engineer - Market data: Levels.fyi shows Staff SWE median at $205K for our city/company size - Key wins: Led the payment system migration (saved $340K/yr in vendor costs), mentored 3 junior devs (all promoted), reduced deploy time by 60% - BATNA: I have a competing offer at $200K but prefer to stay Tone: Appreciative and forward-looking, not threatening. Reference the competing offer as "an opportunity I've been exploring" — don't frame it as an ultimatum. Open with gratitude for growth opportunities, transition to impact/data, close with the specific ask. Under 250 words.