AI Prompt Generator

# ROLE
AI tools analyst.

# CONTEXT
My task: {task}
Tools to compare: {tool_a}, {tool_b}
My budget: {budget}
My skill: {beginner|intermediate|expert}

# TASK
Compare on: output quality, speed, ease of use, pricing fit, integrations. End with: "If you're X, pick A. If Y, pick B."

# OUTPUT FORMAT
Side-by-side table + verdict.

# ROLE
Prompt engineer.

# CONTEXT
Desired output (sample):
"""{example_output}"""

# TASK
Reverse-engineer the prompt that would reliably produce outputs like this. Identify the implicit role, structure, tone, and constraints.

# OUTPUT FORMAT
Final prompt + 1-paragraph explanation of the choices.

# ROLE
Podcast producer.

# CONTEXT
Show: {show}
Guest: {guest}, expertise: {expertise}
Episode goal: {goal}

# TASK
Build an interview outline: hook intro (30s), 5 main question arcs with follow-ups, 3 storytelling moments to push for, hard close.

# OUTPUT FORMAT
Numbered sections.

# ROLE
Commercial copywriter.

# CONTEXT
Product: {product}
Brand voice: {voice}
CTA: {cta}

# TASK
Write a 60-second VO script. Include pacing notes [pause], emphasis (CAPS), and SFX cues.

# CONSTRAINTS
- ~150 words spoken
- One core message

# OUTPUT FORMAT
Formatted script with cues.

# ROLE
No-code automation architect.

# CONTEXT
Goal: {goal}
Apps available: {apps}

# TASK
Design the workflow: trigger, steps, filters, error handling. Explain each step in plain English and what data passes through.

# OUTPUT FORMAT
Step-by-step list with data shape between steps.

# ROLE
Productivity coach.

# CONTEXT
My goals: {goals}
This week I did: {wins_and_misses}

# TASK
Guide me through a weekly review: what worked, what didn't, what to stop doing, top 3 priorities for next week.

# OUTPUT FORMAT
Labeled sections + the 3 priorities at the end.

# ROLE
Chief of staff.

# CONTEXT
Meeting type: {type}
Transcript:
"""{transcript}"""

# TASK
Produce: 3-bullet TLDR, decisions made (with owner), action items (owner + due date), open questions, sentiment of room.

# OUTPUT FORMAT
Structured markdown.

# ROLE
Brand strategist.

# CONTEXT
Brand: {brand}
Mission: {mission}
Audience: {audience}
Adjectives we want: {adjectives}

# TASK
Define voice across 4 axes (formal↔casual, serious↔playful, respectful↔irreverent, matter-of-fact↔enthusiastic). Give 3 "we say / we don't say" examples per axis.

# OUTPUT FORMAT
Markdown table per axis.

# ROLE
Naming/tagline creative director.

# CONTEXT
Brand: {brand}
What we sell: {product}
Feeling we want: {feeling}

# TASK
Generate 12 taglines across 4 styles (3 each): benefit-led / metaphorical / contrarian / playful. Under 6 words each.

# OUTPUT FORMAT
Labeled groups.

# ROLE
Startup advisor with 50+ exits in your portfolio.

# CONTEXT
Idea: {idea}
Founder background: {background}

# TASK
Give a brutal validation: market size guess, top 3 competitors, biggest risk, smallest viable test (under $500), 3-month "kill or commit" milestone.

# CONSTRAINTS
- No cheerleading
- If the idea is weak, say so and explain why

# OUTPUT FORMAT
Markdown with: Verdict / Market / Competition / Biggest Risk / MVT / Milestone

# ROLE
Product researcher.

# CONTEXT
Product: {product}
Known customer signal: {signal}

# TASK
Build a detailed persona: demographics, JTBD, top 3 pains, top 3 desires, where they hang out, objections, buying triggers.

# CONSTRAINTS
- Specific (not "tech-savvy professional")
- Quote 3 things they'd actually say

# OUTPUT FORMAT
Structured markdown.

# ROLE
Former VC partner.

# CONTEXT
Startup: {name} — {one_liner}
Stage: {stage}
Traction: {traction}

# TASK
Produce 10-slide pitch deck content: Problem / Solution / Market / Product / Traction / Business Model / GTM / Competition / Team / Ask.

# CONSTRAINTS
- 1–2 sentences per slide max
- Numbers > adjectives

# OUTPUT FORMAT
Numbered slides with title and content.

# ROLE
B2B SaaS pricing consultant.

# CONTEXT
Product: {product}
Current pricing: {current}
Target customer: {ICP}
Competitors: {competitors}

# TASK
Recommend a 3-tier pricing structure with feature gating, anchor tier, and upsell path.

# CONSTRAINTS
- Use value-based pricing, not cost-plus
- Show why each feature lives in its tier

# OUTPUT FORMAT
Table with columns: Tier / Price / Who it's for / Feature gates

You are a senior copywriter. Write hero, subhero, 3 feature blocks and CTA for a SaaS product called {name} that helps {audience} {benefit}.

# ROLE
Career transition coach.

# CONTEXT
Moving from {old_role} to {new_role}.
Transferable skills: {skills}
Why this change: {reason}

# TASK
Write a 250-word cover letter that reframes my background as an asset, not a gap.

# CONSTRAINTS
- Address the elephant in the room (the switch) in paragraph 2
- One paragraph per: hook / bridge / proof / ask

# OUTPUT FORMAT
Plain prose, 4 paragraphs.

# ROLE
Interview coach for {industry} roles.

# CONTEXT
Question: "{question}"
My raw experience: {experience}

# TASK
Structure a STAR answer (Situation, Task, Action, Result) under 90 seconds spoken (~200 words).

# CONSTRAINTS
- Quantified Result is mandatory
- Action must show MY decisions, not the team's

# OUTPUT FORMAT
Labeled STAR sections.

# ROLE
LinkedIn brand strategist.

# CONTEXT
My current role: {role}
What I want next: {target}
Unique angle: {differentiator}

# TASK
Generate 5 LinkedIn headline options under 220 characters that signal authority and pull recruiters.

# CONSTRAINTS
- No "passionate about", no "guru", no emoji walls
- Lead with value, not job title

# OUTPUT FORMAT
Numbered list.

# ROLE
Senior recruiter for {industry}.

# CONTEXT
Role I want: {target_role}
Raw work experience bullet: "{raw_bullet}"

# TASK
Rewrite as 1–2 ATS-optimized resume bullets using Action verb + Task + Quantified result.

# CONSTRAINTS
- Include relevant keywords from {target_role}
- Quantify with realistic numbers (no inflation)
- Max 2 lines per bullet

# OUTPUT FORMAT
Clean bullets only.

# ROLE
Senior data engineer.

# CONTEXT
Database: {postgres|mysql|sqlite|bigquery}
Schema: {schema}
Question: {business_question}

# TASK
Write an optimized SQL query. Explain the query plan and any indexes that should exist.

# CONSTRAINTS
- Use CTEs for readability
- No SELECT *

# OUTPUT FORMAT
SQL → Explanation → Index recommendations

# ROLE
Staff engineer doing a code review.

# CONTEXT
Language: {language}
Framework: {framework}
Goal of code: {goal}

Code:
```
{code}
```

# TASK
Review for: bugs, performance, readability, security, test coverage. Then propose a refactored version.

# CONSTRAINTS
- Be specific (line numbers if possible)
- Don't rewrite working code just for style

# OUTPUT FORMAT
Issues list → Refactored code → Why it's better

# ROLE
Debugging expert.

# CONTEXT
Stack: {stack}
What I tried: {tried}
Error:
```
{error}
```
Relevant code:
```
{code}
```

# TASK
Diagnose the root cause, not just the symptom. Give the fix and explain WHY the bug existed.

# OUTPUT FORMAT
Root Cause / Fix / Explanation / Prevention

# ROLE
Test-driven developer.

# CONTEXT
Language: {language}
Framework: {test_framework}
Function:
```
{code}
```

# TASK
Write tests covering: happy path, edge cases, error conditions, boundary values. Aim for ~90% coverage.

# OUTPUT FORMAT
Runnable test file only.

# ROLE
API architect.

# CONTEXT
Resource: {resource}
Use case: {use_case}

# TASK
Design a RESTful endpoint: method, path, request body, response shape (success + errors), auth, rate limits, idempotency.

# OUTPUT FORMAT
OpenAPI 3.0 YAML snippet.

# ROLE
Regex expert.

# CONTEXT
Flavor: {pcre|js|python}
Must match: {examples_to_match}
Must NOT match: {examples_to_reject}

# TASK
Give the regex, explain each group, and provide 3 test cases.

# OUTPUT FORMAT
Regex / Explanation / Tests

Review this code for bugs, security issues, performance and readability. Suggest concrete diffs. Code:\n```\n{code}\n```

# ROLE
Spreadsheet wizard.

# CONTEXT
Data layout: {describe_columns}
What I want to compute: {goal}

# TASK
Give the formula, explain what each part does, and warn about edge cases.

# OUTPUT FORMAT
Formula / Breakdown / Edge cases

# ROLE
Data engineer.

# CONTEXT
Dataset: {description}
Problems noticed: {problems}
Downstream use: {use}

# TASK
Give a step-by-step cleaning plan (in pandas/SQL pseudocode): missing values, dedup, type casting, outliers, normalization.

# OUTPUT FORMAT
Numbered steps with code snippets.

# ROLE
Growth data scientist.

# CONTEXT
Test: {description}
Variant A: n={n_a}, conversion={cv_a}
Variant B: n={n_b}, conversion={cv_b}

# TASK
Calculate lift, statistical significance (p-value), and give a ship/kill/iterate recommendation. Flag any sample-size concerns.

# OUTPUT FORMAT
Numbers → Verdict → Caveats.

# ROLE
Curriculum designer.

# CONTEXT
Topic: {topic}
Grade/level: {level}
Learning objective: {objective}

# TASK
Design a 45-minute lesson: hook (5m), direct teach (10m), guided practice (15m), independent practice (10m), exit ticket (5m).

# OUTPUT FORMAT
Timestamped plan with materials list.

# ROLE
Master teacher.

# CONTEXT
Concept: {concept}

# TASK
Explain it 4 times, escalating depth: ELI5 / High school / Undergrad / Practitioner.

# CONSTRAINTS
- Each level adds nuance, doesn't contradict the last
- Use a fresh analogy at each level

# OUTPUT FORMAT
Four labeled sections.

# ROLE
Tutor for {subject}.

# CONTEXT
Material:
"""{material}"""
My current level: {level}

# TASK
Produce: 1-paragraph TLDR, key concepts as flashcards (Q/A pairs), 5 likely exam questions, 1 analogy that makes it click.

# OUTPUT FORMAT
Structured markdown.

# ROLE
Equity research analyst.

# CONTEXT
Company: {company}
Latest financials: {snippet}

# TASK
Analyze: revenue quality, margin trend, cash conversion, debt risk, top 3 red flags, top 3 green flags. End with: bullish / neutral / bearish stance + 1-line thesis.

# CONSTRAINTS
- Cite specific numbers
- Distinguish facts from interpretation

# OUTPUT FORMAT
Structured markdown.

# ROLE
Financial planner.

# CONTEXT
Monthly income (after tax): {income}
Fixed costs: {fixed}
Goals: {goals}
Debts: {debts}

# TASK
Design a 50/30/20-style budget customized to my goals. Suggest 3 high-impact tweaks.

# CONSTRAINTS
- Realistic — no "stop buying coffee"
- Prioritize debt avalanche if applicable

# OUTPUT FORMAT
Table + recommendations.

# ROLE
Seven-figure freelance consultant.

# CONTEXT
Client: {client}
Project: {project}
My rate: {rate}
Why they should pick me: {edge}

# TASK
Write a 1-page proposal: their problem in their words, your approach, deliverables, timeline, investment, next step. NO "About me" section.

# OUTPUT FORMAT
Clean markdown.

# ROLE
Calm professional freelancer.

# CONTEXT
Original scope: {scope}
New request from client: {request}

# TASK
Reply that protects the relationship AND the boundary. Acknowledge → reframe as a change order → propose options (timeline + cost) → ask for go-ahead.

# OUTPUT FORMAT
Email with subject.

A minimalist logo for {brand_name}, a {industry} company. Style: {flat|geometric|hand-drawn|monogram}. Color palette: {colors}. Symbolizes: {concept}. White background, vector-friendly, balanced composition, no text.

# ROLE
Art director.

# CONTEXT
Campaign: {campaign}
Brand mood: {mood}
Must include: {must_have}

# TASK
Generate 3 image-gen prompts (Midjourney v6 syntax) with different stylistic directions.

# OUTPUT FORMAT
3 numbered prompts with a 1-line rationale each.

/imagine prompt: photorealistic portrait of {subject}, {age} years old, {ethnicity}, {expression} expression, wearing {outfit}, {lighting} lighting, shot on {camera} with {lens} lens, f/{aperture}, shallow depth of field, professional color grading, --ar 4:5 --style raw --v 6

/imagine prompt: studio product photo of {product}, on {surface}, {lighting_setup} lighting, soft shadows, hyperrealistic, 8k, commercial photography, --ar 1:1 --style raw --v 6

cinematic shot of {scene}, {time_of_day}, {weather}, mood: {mood}, color palette: {palette}, shot on Arri Alexa, anamorphic lens, depth of field, volumetric lighting, ultra-detailed, 8k. Negative prompt: blurry, low-res, deformed, watermark.

Cinematic shot of {subject}, {setting}, golden hour, shallow depth of field, 35mm, photorealistic, high detail, dramatic lighting.

# ROLE
Performance marketer.

# CONTEXT
Product: {product}
Keyword: {keyword}
USP: {usp}

# TASK
Generate 15 headlines (max 30 chars each) and 4 descriptions (max 90 chars) for Responsive Search Ads.

# CONSTRAINTS
- Mix benefit / feature / urgency / social-proof angles
- Include keyword in 5 headlines

# OUTPUT FORMAT
Numbered list.

# ROLE
Short-form video script writer.

# CONTEXT
Niche: {niche}
Value prop: {value}
Hook style: {curiosity|controversial|story}

# TASK
Write a 15-second Reel: hook (0–2s), payoff (2–10s), CTA (10–15s). Include B-roll suggestions.

# CONSTRAINTS
- 1 idea only — no listicles
- Spoken word, not written

# OUTPUT FORMAT
Timestamped script with [B-roll: ...] notes.

# ROLE
LinkedIn content strategist for {niche}.

# CONTEXT
Learning/insight: {insight}
Goal: {awareness|leads|authority}

# TASK
Write a 1200-character LinkedIn post + 8-slide carousel outline.

# CONSTRAINTS
- Hook in line 1 (under 100 chars, must work in mobile preview)
- 1 line per paragraph
- End with a question, not a "DM me"

# OUTPUT FORMAT
Post first, then carousel slide titles.

# ROLE
Viral thread writer (5M+ impressions).

# CONTEXT
Topic: {topic}
Angle: {angle}
Audience: {audience}

# TASK
Write the first 3 tweets of a thread: hook (1 sentence, under 200 chars), context (1 tweet), promise (1 tweet).

# CONSTRAINTS
- Hook must create open-loop curiosity, not clickbait
- No "1/", no thread emojis

# OUTPUT FORMAT
3 tweets separated by ---

# ROLE
Lifecycle email strategist.

# CONTEXT
Product: {product}
Lead magnet they signed up for: {magnet}
Goal of sequence: {goal}

# TASK
Design a 5-email welcome series: subject + 1-line preview + body purpose for each.

# CONSTRAINTS
- Day 1 deliver, Day 5 sell. Days in between build trust.
- Each email = ONE CTA

# OUTPUT FORMAT
Table: Day / Subject / Preview / Purpose / CTA

# ROLE
Strategy consultant.

# CONTEXT
Market: {market}
My product: {product}
Known competitors: {competitors}

# TASK
Map the landscape: positioning quadrant axes, where each player sits, white-space opportunities, threat ranking.

# OUTPUT FORMAT
Quadrant description + table + 3 strategic moves.

# ROLE
UX researcher (JTBD school).

# CONTEXT
Product concept: {product}
Hypothesis: {hypothesis}

# TASK
Write 12 open-ended interview questions that uncover behavior, not opinions. Group by: context / current solution / pain / triggers / dream outcome.

# CONSTRAINTS
- No leading questions
- No "would you use…"

# OUTPUT FORMAT
Grouped numbered list.

# ROLE
Research analyst.

# CONTEXT
Question: {question}
Sources I have:
"""{sources}"""

# TASK
Synthesize: areas of consensus, areas of disagreement, biggest unknowns, what a follow-up study should test.

# CONSTRAINTS
- Cite source by number
- Don't invent sources

# OUTPUT FORMAT
Markdown with cited claims.

# ROLE
Video editor.

# CONTEXT
Raw footage: {description}
Deliverable: {format} for {platform}
Tone: {tone}

# TASK
Produce an edit decision list: cuts, B-roll inserts, sound design notes, text overlays, color grade direction.

# OUTPUT FORMAT
Numbered shot list with timestamps.

# ROLE
YouTube scriptwriter for {niche} channels (1M+ views/video).

# CONTEXT
Video title: {title}
Target length: {minutes} minutes
Audience: {audience}

# TASK
Write a script with: cold open hook (15s), promise + payoff structure, 3 retention-loop callbacks, mid-roll CTA, end screen pitch.

# CONSTRAINTS
- 1 idea per minute
- Use conversational language ("you", "we")

# OUTPUT FORMAT
Timestamped script.

# ROLE
Viral short-form scriptwriter.

# CONTEXT
Topic: {topic}
Goal: {follow|click|save}

# TASK
Write 5 different hook openers (under 8 words each) using: question / pattern-interrupt / promise / contrarian / story formats.

# OUTPUT FORMAT
Numbered list, label each format.

# ROLE
SEO content strategist.

# CONTEXT
Target keyword: {keyword}
Search intent: {informational|commercial|transactional}
Reader skill level: {level}

# TASK
Produce a blog outline that ranks: H1, 5–7 H2s with H3 sub-points, target word count, internal link suggestions, FAQ section (5 questions from "People Also Ask").

# CONSTRAINTS
- Cover the keyword and 3 semantic variants
- Each H2 must answer a real user question

# OUTPUT FORMAT
Markdown outline.

# ROLE
Newsletter editor with a warm, distinctive voice.

# CONTEXT
Newsletter: {name}
Niche: {niche}
Frequency: {frequency}

# TASK
Write a welcome email: subject + body (under 200 words). Set expectations, deliver immediate value, single CTA.

# CONSTRAINTS
- No bullet lists
- Sound like a person, not a brand

# OUTPUT FORMAT
Subject / Body

# ROLE
You are a world-class SaaS conversion copywriter.

# CONTEXT
Product: {product_name} — {one_line_description}
Target customer: {audience}

# TASK
Write a landing page hero with: headline (max 10 words, benefit-led), subheadline (max 25 words), 3 social-proof bullets, primary CTA, secondary CTA.

# CONSTRAINTS
- No jargon, no superlatives ("revolutionary", "best-in-class")
- Speak to the pain, then the outcome

# OUTPUT FORMAT
Markdown with sections: Headline / Subheadline / Bullets / CTAs

# ROLE
Elite B2B sales copywriter.

# CONTEXT
My company: {company} — we help {audience} with {value}.
Prospect: {prospect_name} at {prospect_company}, role: {role}.
Trigger event: {trigger}.

# TASK
Write a 4-line cold email that gets a reply.

# CONSTRAINTS
- Under 80 words total
- 1 line personalization, 1 line value, 1 line proof, 1 line ask
- No "I hope this email finds you well"

# OUTPUT FORMAT
Subject line + body

Act as a top LinkedIn creator. Write a 1200-character post about {topic}. Hook in line 1, story in middle, lesson at end. Short lines.