Prompt library

System: Be concise. No preamble. No filler phrases.
Answer directly. Use markdown bullets for lists.
Max 150 words unless explicitly asked for more.
If clarification needed, ask ONE specific question only.

Custom Instructions (What would you like ChatGPT to know about you):
I'm a [your role] working on [your domain]. Be direct, skip disclaimers.

Custom Instructions (How should ChatGPT respond):
No preamble. Bullet points preferred. Max 200 words unless asked more. Skip apologies.

System: You are a [role] assistant for [context].
Respond in markdown. Be concise, max 200 words.
Use Google Search when needed for current information.
No disclaimers or apologies.

System: You are a fast, direct assistant.
Answer in under 100 words unless more is needed.
Use bullet points for lists. No disclaimers.

Research task using X/Twitter data:
Topic: [brand/product/topic name]
Find: [trending discussions / sentiment / top complaints / viral posts]
Time frame: [last 24h / last week]
Output: [summary / sentiment score / top 5 posts / actionable insights]
Context: I'm a [your role] tracking [your purpose].

For structured data extraction (Mistral JSON mode):
Extract the following fields from the document below:
- [field_1: type and format]
- [field_2: type and format]
- [field_3: list of items]
Return ONLY valid JSON with no additional text.
Document: [paste document content]

.cursorrules template:
Project: [project name]
Stack: [React 18, TypeScript, Tailwind CSS, Next.js 14 App Router]
Style: [functional components, arrow functions, no class components]
Naming: [PascalCase components, camelCase functions, UPPER_SNAKE constants]
Do not: [use any, use var, use inline styles]
Always: [add TypeScript types, handle loading/error states, use shadcn/ui components]

// Task: [what this function/component should do]
// Input: [parameters and their types]
// Output: [return value and type]
// Edge cases: [null handling, validation needed]
// Dependencies: [libraries/functions it should use]
function

.windsurfrules example:
You are working on a [Next.js 14 + TypeScript + Tailwind] project.
Always use: App Router, Server Components by default, shadcn/ui components.
Never use: class components, any type in TypeScript, inline styles.
When creating components: add loading states, error boundaries, and mobile responsive design.
File naming: PascalCase for components, camelCase for utils, kebab-case for routes.

# CLAUDE.md
## Project: [Name]
## Architecture
[Brief description of how the project is organized]
## Key files
- [path]: [what it does]
## Commands
- `npm dev`: start dev server
- `npm test`: run tests
## My coding standards
[list your preferences]
## What to avoid
[antipatterns and things that have caused bugs]

[Subject description], [artistic style or medium], [lighting], [mood/atmosphere], [specific details], photographed by [photographer style] --ar 16:9 --v 6.1 --s 200 --style raw

A [subject], [art style/medium], [lighting conditions], [color palette], [mood], [camera angle if applicable]. The image should feel [adjective]. Include [specific element]. --No text overlay unless specified.

[POSITIVE] masterpiece, best quality, ultra-detailed, [subject], [style], [lighting], [camera]
[NEGATIVE] lowres, bad anatomy, bad hands, text, error, missing fingers, extra digit, fewer digits, cropped, worst quality, low quality, jpeg artifacts, watermark, blurry

[Subject with exact text you need in quotes, e.g. a modern logo with the word "Maven"], design style, [color palette: primary #00e5ff, secondary #07070c], clean, minimal, vector, white background --style design --ar 1:1

Figma AI prompt format (Make Designs):
[Screen type: Dashboard / Landing page / Mobile settings / Checkout]
Include: [navigation bar, sidebar with stats, data table, action buttons]
Style: [dark UI with blue accents / minimal white / glassmorphism]
Elements: [show user profile, notifications bell, search bar]
Layout: [desktop 1440px / mobile 375px / tablet 768px]

AI Layout: Create a [page type, e.g. 'pricing page'] for a [product type].
Target user: [audience]
Key sections: [hero, features, pricing, CTA]
Design style: [modern dark / clean light / minimal]
Primary color: [hex]
Include: [specific elements]

Autodesigner prompt: [App name] — [one sentence description]
Type: [web app / iOS app / Android app]
Key screens: [list 4–6 main screens]
Style: [modern, minimal, colorful, professional]
Primary color: [hex or color name]
Target user: [audience in 1 sentence]

Text to UI: [Screen name] for [product type]
Purpose: [what this screen does]
Key elements: [navigation, hero section, list/grid, CTA button, form, etc.]
Layout: [mobile/tablet/desktop]
Style: [minimal/detailed/dark/light]
Important: include [specific UI element that must appear]

Build a [specific feature name] for my [app type].
Existing context: [what already works in the app]
This feature should: [3–5 specific bullet points of what it does]
Do NOT change: [list components that are working]
Use: [existing component names if applicable]
Expected result: [describe the end state precisely]

Build a [app/feature description] using React, TypeScript, Vite, and Tailwind CSS.
Features:
- [Feature 1 with specific behavior]
- [Feature 2]
- [Feature 3]
Style: dark theme, [primary color], Outfit font, mobile-responsive.
Do not use: [libraries or patterns to avoid]
Start with: [the most important core feature first]

Research question: [specific question with context]
Focus: [Academic / Reddit / News / YouTube / Web]
I need: [specific type of answer — overview / deep dive / comparison / latest updates]
Format: [bullet points / paragraph / table]
Time range: [if relevant — last 6 months / 2026 only]

For best TTS results, write: Short sentences. Active voice. Avoid abbreviations — write them out (e.g. 'Doctor' not 'Dr.'). Add commas where you want natural pauses. Write numbers as words ('twenty-five thousand' not '25,000'). Use em dashes — like this — for dramatic pauses.

Style: [genre], [tempo: fast/medium/slow], [mood: upbeat/melancholic/energetic], [instruments: acoustic guitar, piano, drums], [vocal style: female pop, male indie, no vocals]
[Verse 1]
[your lyrics here]
[Chorus]
[chorus lyrics]
[Bridge]
[bridge lyrics]

[Scene description], [subject action: a woman walks slowly], [environment: neon-lit Tokyo street, rain, night], [camera: slow zoom out, static, handheld shakiness], cinematic, film grain, 24fps, anamorphic, shallow depth of field --duration 5s

HeyGen Video Script Format:
[Avatar action: looking at camera / gesturing left]
[Script: clear sentences, 130-150 words/minute pace]
[Pause: add '...' for 0.5s pause, '....' for 1s pause]
[Emphasis: CAPITALIZE words you want stressed]
Keep total video under 3 minutes for best engagement.

Synthesia Script Format:
Slide 1 (Title): [Topic Name]
Avatar says: Welcome to [course name]. Today we'll cover [3 key points].
[New Slide]
Avatar says: [Keep sentences under 20 words. Pause naturally. Avoid jargon.]
On-screen text: [Key term or stat shown simultaneously]
Duration target: [aim for 130 words per minute = 3 min for 390 words]

[Subject description and action], [environment and time of day], [camera motion: slow dolly in, tracking shot, aerial descent], [lighting: golden hour, studio softbox, neon night], [style: cinematic, documentary, music video aesthetic], realistic, photorealistic, 24fps

[Subject description], [Pikaffect if desired: 'dissolve into liquid', 'inflate like balloon'], [environment], [motion: floating upward, spinning slowly], [lighting and mood], [camera: static close-up, slow pan], short social video style

Create a [component name] React component with:
- [Key feature 1 with specific behavior]
- [Key feature 2]
- [Visual style: dark theme / glassmorphism / minimal]
Use shadcn/ui components where possible.
Make it: responsive (mobile-first), accessible (ARIA labels), dark mode compatible.
Props: [list the component's configurable props]

Replit Agent prompt format:
Build a [app type] that [core functionality in one sentence].
Users can: [action 1], [action 2], [action 3]
Data: [what data is stored — use a simple JSON file or Replit DB]
Style: [clean, minimal | material | custom: dark with blue accents]
Deploy it at the end.
Do not use: [technologies to avoid]

Firefly works best with Adobe-style descriptors: '[main subject], [Adobe Stock photo style], [composition: rule of thirds, centered, aerial view], [lighting: soft studio, dramatic side-light], [color palette: warm earth tones, high contrast], commercial photography quality, [output use: web banner, social post, print ad]'

Magic Design prompt: [content topic], [target audience], [tone: professional/playful/bold], [specific elements to include: product image, CTA button, testimonial], [color preference: use brand colors / warm / cool / vibrant]

Magic Write prompt: Write a [type: headline/tagline/caption/CTA] for [product/service] targeting [audience] in a [tone] tone. Max [word count] words.

Marketing copy brief:
Product: [name and one-line description]
Audience: [specific persona: Indian startup founders aged 25-35]
Goal: [awareness / conversion / retention]
Channel: [email subject line / Facebook ad / landing page hero]
Tone: [professional, conversational, urgent]
CTA: [sign up free / book a demo / buy now]
Unique angle: [what makes this different from competitors]
Word limit: [50 words for ad / 150 for email / 300 for landing page]

Grammarly GO prompt for email drafting:
Context: I need to [email goal: decline a meeting / follow up on proposal / introduce myself]
Recipient: [name, title, relationship — existing client / new prospect / senior manager]
Key points to cover: [point 1], [point 2], [point 3]
Tone: [professional and warm / direct and brief / apologetic but firm]
Length: [3 sentences / 1 paragraph / full formal email]
Avoid: [jargon / aggressive language / excessive apologies]

Notion AI task prompts:
[Summarize this page into 5 bullet points]
[Turn these meeting notes into action items with owners and due dates]
[Write a PRD for this feature idea based on the notes below]
[Find all mentions of [topic] across my workspace and summarize them]
[Draft an email to stakeholders summarizing this project update]

Murf voiceover script format:
[Opening hook — short, punchy, first 5 seconds]
[Problem statement — empathy with audience]
[Solution: your product/service name in CAPS for emphasis]
[Key benefits: 3 points, one sentence each]
[CTA: clear, specific next step]
[Add pauses: ... for 0.5s, use [P1] for 1s before important points]
Target duration: [60s = ~150 words | 90s = ~225 words | 2min = ~300 words]

Descript Underlord AI prompt for social clips:
From this [podcast episode / course video / interview], find:
- [3 most engaging moments for social media clips]
- [The key insight or takeaway for the episode description]
- [5 timestamps for chapter markers]
- [Hook sentence for YouTube title]
Format each clip with: [timestamp], [subject], [why it's engaging]

Surfer SEO + AI Writer brief:
Keyword: [primary keyword]
Secondary keywords: [keyword 2], [keyword 3]
Search intent: [informational / transactional / navigational]
Target word count: [as recommended by Surfer for this keyword]
Heading structure: [H1 once, H2 for main sections, H3 for sub-points]
Include: [specific questions from People Also Ask section]
Content Score target: [70+ minimum / 80+ for competitive keyword]

GTM Workflow prompt:
Task: Generate personalized cold email for [prospect name] at [company]
Input data: Company = [name], Industry = [industry], Pain = [known pain point], Our solution = [what you offer]
Email structure:
- Subject line: curiosity-based, under 50 chars
- Opening: personalized observation about their business
- Problem: their likely pain point in one sentence
- Solution: your product in one line, no pitch language
- CTA: one specific, low-friction ask
Tone: direct, conversational, peer-to-peer (not sales-y)

[Subject description], [specific style element using Elements LoRA names], [lighting: dramatic rim light / soft studio / golden hour backlight], [composition: rule of thirds, close-up portrait, wide environmental shot], [quality tags: highly detailed, 8K, professional photography, masterpiece] --model photorealistic --guidance 7

InVideo AI video brief:
Topic: [specific topic, not generic — 'How to file GST returns in India 2026']
Target: [YouTube explainer for first-time business owners]
Tone: [simple, friendly, step-by-step]
Duration: [5 minutes]
Include: [real screen recordings if available / infographics for key data / text on screen for every key point]
Voiceover language: [English / Hindi]
Add captions: [Yes — required for silent viewing]

Meeting transcription best practice:
Task: Record and summarize a [client call / internal standup / sales demo]
Automatic actions:
1. Fireflies records audio
2. AI generates transcript with speaker labels
3. Action items extracted with owner assignments
4. Summary generated (bullet points)
5. Integrated into CRM (HubSpot) with call notes linked
Output format:
- Timestamp + speaker name + statement
- Key discussion points highlighted
- Decisions made with owner/deadline
- Next steps explicitly listed

Otter meeting workflow:
Task: Transcribe and summarize [type of meeting]
Setup:
1. Add Otter Assistant to Zoom/Teams meeting link
2. Meeting starts, Otter joins and transcribes in real-time
3. Attendees see live captions in meeting window
4. Speakers mark highlights ('highlight this') as they speak
Post-meeting:
- Full transcript with timestamps
- Auto-summary: decisions, action items, next steps
- Search for specific topics across all past meetings
Output: Shareable transcript link + PDF download

Mem daily note template:
# [Date] — [Project Name]

WHAT I DID:
- [accomplishment 1]
- [accomplishment 2]

WHAT I LEARNED:
- [insight 1]
- [decision made]

BLOCKING ME:
- [issue 1]

NEXT STEPS:
- [tomorrow's priority 1]

#project #learning

Readwise reading system:
Goal: Capture knowledge from 20+ sources without information overload

Daily routine:
1. Morning: Skim Readwise Reader inbox (3–5 min)
2. Highlight key passages (3–4 per article read)
3. Readwise auto-syncs to Obsidian with tags
4. Evening: Spaced rep review of old highlights (2 min)

Weekly: Review topics in Readwise — which themes emerged? Add to knowledge base.

NotebookLM research workflow:
Task: Synthesize knowledge from [10 articles / 5 research papers / company docs] on [topic]

Steps:
1. Upload all sources to single notebook
2. Ask clarifying questions: 'What are the 3 main disagreements between these sources on [topic]?'
3. Generate Audio Overview (podcast) to understand core ideas
4. Ask follow-ups: 'Which source provides the strongest evidence?'
5. Generate Study Guide for sharing with team
6. Export as structured document

Output: Synthesis with citations, audio summary, FAQ guide

Elicit research workflow:
Task: Synthesize evidence on [research question]

Prompt to Elicit:
'[Specific research question in detail]
Please extract:
1. Study design (RCT, observational, etc.)
2. Sample size and population
3. Main result/effect size
4. Year published
Focus on studies from [date range / region / population]'

Output: Structured table ready for meta-analysis or evidence synthesis

Consensus research validation:
Task: Find evidence quality for [specific claim]

Query to Consensus:
'[Research question or claim to validate]
Show me:
1. Percentage consensus (do most studies agree?)
2. Top papers supporting this
3. Any contradictory findings
4. Study quality (RCT vs. observational)'

Output: Evidence rating (Strong | Moderate | Mixed | Limited), key papers, consensus percentage

Character creation template:
NAME: [Character name]
BACKGROUND: [5–10 sentence biography]
PERSONALITY: [3–4 key traits: curious, sarcastic, analytical, warm]
EXPERTISE: [What they're known for]
SPEAKING STYLE: [Formal / Casual / Poetic / Direct / Verbose]
MOTIVATION: [What drives them in conversation]
EXAMPLE RESPONSE: [Show a typical way they'd answer a question]

Poe bot for market research:
Name: Market Analyst
Base model: Claude 3.5 Sonnet
System prompt: 'You are a market research analyst specializing in India's SaaS and D2C sectors. Analyze data critically. Provide market size, growth rate, competitive landscape, and investment trends. Always cite sources. Flag assumptions.'
Knowledge files: [Upload recent market reports, funding data, industry analyses]
Usage: Ask about specific markets, get analysis, switch to GPT-4o for contrasting perspective

You.com research workflow:
Task: Research [industry trend / competitor analysis / product research]

Search in You.com AI mode:
'[Specific research question]'
Output:
- AI summary of findings
- Top 5 sources cited
- Related questions suggested
Switch to Traditional mode for:
- Specific URLs / blog posts / whitepapers
Add to Collection for:
- Long-term research projects

Phind search for debugging:
Query: '[Error message or problem description]'
Example: 'TypeError: Cannot read property 'map' of undefined'
Phind returns:
- Root cause explanation
- 3–5 solutions ranked by relevance
- Code examples with context
- GitHub issues with same problem
- Related Stack Overflow threads

For new features:
Query: '[I want to build X in Y framework]'
Example: 'Build a real-time comment system with WebSockets in Next.js'
Phind returns: Code skeleton + multiple approaches + performance considerations

Writesonic blog post creation:
Topic: [Your blog topic]
Keyword: [Primary keyword (max 5 words)]
Target audience: [Who is reading this?]
Tone: [Professional / Friendly / Technical]
Length: [Word count: 2000]
Structure:
- Engaging headline
- Hook paragraph
- 4–5 H2 sections with subheadings
- Conclusion with CTA
Include:
- Real examples relevant to India market
- At least 3 statistics with sources
- 1 surprising insight

Rytr template usage:
Use case: Product Description
Product: [Name and 2-line description]
Key features:
- [Feature 1]
- [Feature 2]
- [Feature 3]
Target buyer: [Who is buying this]
Unique angle: [What makes it different from competitors]
Tone: [Casual / Professional]

Rytr generates: Compelling product description ready to copy to your store

QuillBot workflow for academic writing:
1. Write draft in natural voice
2. Paste into QuillBot
3. Use Formal mode → academic tone
4. Use Expand mode → increase word count for requirements
5. Grammar check → catch errors
6. Plagiarism check → verify citations
7. Export revised version

For business writing:
1. Paste rough draft
2. Use Professional mode
3. Adjust tone (Formal vs. Friendly)
4. Check grammar
5. Final review

Wordtune writing workflow:
Step 1: Write naturally (don't worry about perfect phrasing)
Step 2: Highlight problem sentence
Step 3: Open Wordtune
Step 4: Choose tone/style (Formal | Casual | Shorter | Longer | Clearer)
Step 5: Pick best alternative from options
Step 6: Continue writing

ForEmail:
- Write draft in friendly tone
- Use Wordtune to make 'More formal' for client
- Use 'Shorter' to cut unnecessary words
- Keep clarity high

Anyword A/B testing workflow:
Task: Write 5 email subject lines for [product launch]

Variations:
1. '[Subject line 1 — feature-focused]'
2. '[Subject line 2 — benefit-focused]'
3. '[Subject line 3 — urgency-focused]'
4. '[Subject line 4 — curiosity-focused]'
5. '[Subject line 5 — personalization-focused]'

Anyword scores each:
- Predicted open rate
- Strength (power words, length, structure)
- Audience resonance
Send the highest-scoring variation

Gamma presentation generator:
Prompt: 'Create a presentation about [topic] for [audience]
Tone: [Professional/Casual/Urgent]
Main points:
1. [Point 1 with 2-3 sub-bullets]
2. [Point 2 with 2-3 sub-bullets]
3. [Point 3 with 2-3 sub-bullets]
Include sections:
- Opening/Problem statement
- Current state (data-driven)
- Proposed solution
- Next steps
Brand colors: [Color 1] [Color 2] [Color 3]'

Beautiful.AI investor pitch template:
[Slide 1] Title slide: Company name + tagline
[Slide 2] Problem: Customer pain point (with icons/image)
[Slide 3] Solution: How you solve it (with diagram)
[Slide 4] Market size: TAM/SAM/SOM with numbers
[Slide 5] Traction: User growth/MRR/customer wins (with chart)
[Slide 6] Team: Founder profiles (with photos)
[Slide 7] Financials: Revenue and burn (with graph)
[Slide 8] Ask: Funding amount + use of funds

Beautiful.AI auto-designs every slide

Udio music generation:
Prompt: '[Genre] [mood] [instrumentation] song about [topic] with [optional: tempo/length]'

Example 1:
'Upbeat lofi hip-hop instrumental, chill vibe, drums + vinyl crackle, 60 seconds'

Example 2:
'Sad piano ballad with strings, melancholic, 2 minutes, ambient pad underneath'

Example 3:
'Energetic Indian fusion track blending sitar with modern electronic beats, 90 BPM, 2 minutes'

Aftergenerating:
- Download MP3 (copyright-free)
- Remix in different style
- Use for videos/podcasts/streams

Pictory video from article:
Source: [Blog post URL or pasted article]
Topic: [Article topic]
Target: [YouTube / Instagram Reels / TikTok / LinkedIn]
Voice: [Language + gender]
Style: [Professional | Casual | Educational]

Pictory processes:
1. Extracts article text
2. Breaks into scenes
3. Finds matching stock footage
4. Generates voiceover
5. Adds captions
6. Formats for platform
7. Exports video

You can edit at each step

Lumen5 blog-to-video workflow:
Source blog post: [article on [topic]]
Target platform: [Instagram Reels / TikTok / YouTube Shorts / LinkedIn]
Brand voice: [Professional / Casual / Educational / Storytelling]
Clip duration: [15s (short) | 30s (standard) | 1min (long)]
Key messages: [3–5 main points to emphasize]

Lumen5 generates:
- Auto-divided scenes from article
- AI images matching each scene
- Text overlays with key points
- Voiceover + background music
- Platform-optimized dimensions

Kapwing workflow for social content:
Task: Edit [video/image] for [platform]

1. Upload media
2. Use preset dimensions for platform
3. Add captions (auto-caption AI if video)
4. Add text overlays with key messages
5. Add background music (from Kapwing's library or upload)
6. Add stickers/effects for visual interest
7. Trim to platform requirements (TikTok 9:16, YouTube 16:9)
8. Export and download

Alternative: Create meme → upload template → add text → export

Julius data analysis example:
Task: Analyze sales performance
Data: [Sales CSV with Date, Region, Product, Revenue columns]
Questions:
'1. Which region has highest growth trend in Q1-Q2?'
'2. Top 3 products by revenue?'
'3. Is revenue correlated with seasonality?'
'4. Forecast revenue for Q3 assuming growth continues'
'5. Create a dashboard showing all above'

Julius outputs:
- Charts for each question
- Key findings highlighted
- Statistical significance tests
- Forecast with confidence interval

Scite research workflow:
Task: Write literature review on [topic]

1. Input research question into Scite
2. Get list of papers with citation analysis
3. View Smart Citations:
   - Green = supports this claim
   - Red = contradicts this claim
   - Gray = mentions it
4. Build narrative: introduce claim, show supporting papers, acknowledge contradictions
5. Use Cite-While-You-Write in Google Docs
6. Export citations in academic format
7. Scite generates structured review outline

Tabnine setup workflow:
1. Install Tabnine extension in your IDE
2. Tabnine indexes your project (one-time)
3. Start coding, Tabnine suggests completions

Example:
Type: 'def calculate_'
Tabnine suggests: 'calculate_total_price(items: List[Dict]) -> float:'

Press Tab to accept, keep typing to modify

Type: 'fetch_user'
Tabnine suggests full function with error handling, based on your project patterns

You are my professional AI assistant integrated into Microsoft 365.
For documents: be concise, use professional language, match my existing tone.
For data analysis: show formulas and explain the logic.
For emails: match my communication style — [brief/detailed/formal/casual].
Always: structure outputs with headings, avoid filler phrases.

System: You are a precise technical assistant.
For reasoning tasks: show your work step by step inside <think> tags, then give the final answer.
For code: include error handling, type hints (Python), and brief inline comments.
For analysis: structure with numbered conclusions, cite your reasoning.
Be concise — no filler phrases or preamble.

Keep responses brief and conversational — I'm often using you on mobile in WhatsApp.
For tasks: give me the direct answer first, then explanation if needed.
For India-specific topics: include Indian context, pricing in INR, and local examples.
No formal headings for casual questions — just natural conversation.

System: Answer only from the provided documents. If the answer is not in the documents, say 'This information is not in the provided context.'
Cite the source document for every factual claim using [Doc N] notation.
Be concise — max 200 words unless the question requires more detail.
Do not add information from your training data.

# Modelfile example for custom assistant
FROM llama3.3
SYSTEM """
You are a concise, expert AI assistant.
Answer directly without preamble.
For code: add comments and error handling.
For analysis: use bullet points.
Context: Indian user, prefer INR pricing, local examples.
"""
PARAMETER temperature 0.7
PARAMETER num_ctx 8192

[Detailed subject description], [exact lighting: soft diffused studio light from left], [camera: Canon EOS R5, 85mm f/1.4, shallow depth of field], [environment: specific location with details], [mood/atmosphere], [style: photorealistic, documentary, editorial photography], [specific colours: warm amber tones, navy shadows], [output: product photography / portrait / landscape / architectural]

[Subject: specific description], [style: graphic design poster, digital illustration, photorealistic], [colour palette: primary #hex, secondary #hex, background #hex], [composition: centered, rule of thirds, close-up], [additional elements: texture, pattern, typography], [mood: energetic, minimal, luxury], [output use: social media post, website banner, product label]

For Replace Background: '[Product description] on [new background: rustic wooden table, white studio, outdoor Indian market stall], [lighting matching original image direction], photorealistic product photography, [specific mood or style]'

Real-time canvas approach: start with rough shapes in neutral colours → describe the target style in text → adjust style strength slider while Krea renders live → fine-tune composition → upscale final to 4K. More effective than prompt-only generation for creative concepts.

Looka brand brief: Company name: [name], Industry: [specific niche], Target audience: [demographics], Brand personality: [2–3 adjectives: modern/bold/friendly], Colours to use: [if any], Colours to avoid: [if any], Similar brands I like: [2–3 examples], Logo style: [minimal icon / wordmark / combination mark / emblem]

Framer AI site brief: Business: [company name and one-line description], Target audience: [specific persona], Tone: [professional/friendly/bold/minimal], Key pages: [Home, About, Pricing, Contact], Design style: [modern dark / clean white / gradient / glassmorphism], Brand colours: [hex codes], Important CTA: [what you want visitors to do], Inspiration sites: [2–3 URLs]

Webflow CMS blog brief: Collection name: [Blog Posts], Fields: Title (text), Slug (auto), Published Date (date), Author (reference), Category (reference), Featured Image (image), Body (rich text), Summary (text, 160 chars), SEO Title (text), SEO Description (text). Template layout: [header with breadcrumb, featured image hero, body with sidebar TOC, related posts grid].

Zapier automation brief: Trigger: [When X happens in App A], Conditions: [Only if Y is true], Actions: [1. Do this in App B, 2. Then this in App C, 3. Send notification to Slack]. Data needed: [field names from trigger to pass to actions]. Error notification: [email/Slack for failures]. Task estimate: [expected runs per month × steps per run].

Make scenario brief: Trigger: [Webhook / Schedule / App trigger], Input data: [what fields arrive], Processing steps: [list transformations needed], AI step: [classify / summarise / generate content using which model], Output: [send to which apps with which fields], Error handling: [email alert / retry / log to spreadsheet], Expected volume: [operations/day].

n8n AI Agent workflow: Tools to give the agent: [Web Search, Read from Google Sheets, Send Email, Query Database]. System message: 'You are an AI assistant that helps with [task]. Use the available tools to complete requests. Always verify information before taking action. Report what you did and why.' Memory: [use Window Buffer Memory for conversation continuity]. Model: [GPT-4o or Claude Sonnet for reasoning quality].

Fathom post-meeting prompt (via Ask Fathom): 'Summarise this meeting in 5 bullet points. List all action items with owner and deadline. What questions did the client ask that I haven't answered yet? What commitments did I make? Draft a follow-up email based on the discussion.'

Loom video structure template: Opening (15s): What this video covers and why it matters. Context (30s): The problem or situation. Demo/walkthrough (2–3min): Screen recording with verbal explanation. Key points (30s): Summarise what to remember. Next steps (15s): What you need from the viewer or what happens next. Keep total under 5 minutes — viewers drop off after 5 min on async video.

tl;dv post-call debrief template: 'Summarise the meeting in 3 sentences. List action items with owner and due date. What were the top 2 customer pain points mentioned? What competing products did they mention? What is the agreed next step and timeline? Rate the call quality 1–10 and explain why.'

Use Krisp + professional audio setup: Krisp for noise removal + cardioid microphone (₹3,000–₹5,000 Blue Snowball or HyperX SoloCast) + proper mic positioning (6–8 inches from mouth, side-address) = broadcast-quality audio from any location. This setup makes you sound more professional than 95% of Indian professionals on calls.

Play.ht script formatting for natural speech:
- Write short sentences (under 20 words) — long sentences cause robotic pacing
- Spell out numbers and abbreviations: 'twenty-five thousand' not '25,000', 'Doctor' not 'Dr.'
- Add em-dashes for emphasis pauses: 'The result — surprisingly — was different'
- Avoid parenthetical asides in brackets
- Use active voice: 'We built this' not 'This was built by us'
- Test with Turbo first for pacing, then generate full quality

Aiva composition brief: Genre: [orchestral/ambient/electronic/jazz], Mood: [epic/peaceful/melancholic/uplifting], Duration: [seconds], BPM: [60-160], Key: [major for upbeat/minor for drama], Instrumentation: [strings/brass/piano/synth/full orchestra], Dynamic shape: [builds from quiet to climax / stays consistent / starts big then fades], Use case: [film score / game background / podcast intro / corporate video].

Soundraw music brief: Content type: [YouTube video / podcast intro / ad / presentation], Duration: [seconds], Mood: [2–3 mood descriptors], Energy: [low throughout / builds to climax / peaks at [timestamp]s], Instruments preferred: [piano / guitar / full band / electronic], BPM feel: [slow/medium/fast], Similar music: [reference track or artist for style guidance].

[Scene: describe what's happening, who's in it, what action occurs], [Camera: slow dolly push in / aerial descent / static wide shot / handheld follow], [Environment: specific location with time of day and weather], [Lighting: golden hour rim light / harsh midday sun / soft overcast / neon city night], [Mood/atmosphere], [Style: photorealistic 35mm film / cinematic anamorphic / documentary handheld], [Audio: ambient street noise / orchestral music / silence]. Keep descriptions specific and visual, not narrative.

[Subject: specific person/object description], [Action: exactly what motion occurs — a woman slowly turns her head, water ripples outward from a dropped stone], [Environment: specific location and time of day], [Camera: slow push in / orbit at eye level / rising crane shot / locked static], [Lighting: golden hour from left / studio softbox / moonlit fog], cinematic, photorealistic, 24fps, film grain, shallow depth of field.

CapCut Script-to-Video prompt: Topic: [specific subject], Format: [tutorial/listicle/story/talking-head], Duration: [30–60s], Hook (first 3s): [attention-grabbing opening statement or question], Key points: [3–5 main points with B-roll suggestions], CTA: [follow/comment/save], Style: [fast-paced/calm/energetic], Music mood: [upbeat/motivational/calm].

Opus Clip curation brief: Source content: [podcast episode / webinar / YouTube video URL], Target platform: [TikTok/Reels/Shorts], Target audience: [describe niche], Preferred clip style: [educational nuggets / funny moments / inspiring quotes / hot takes], Clip count: [generate top 10 clips], Brand context: [your name/brand], Include/exclude topics: [specific sections to include or avoid].

Devin task specification template:
Task: [specific feature name]
Acceptance criteria:
- [ ] [Specific behaviour 1]
- [ ] [Specific behaviour 2]
- [ ] [Error case handled]
Existing patterns to follow: [reference specific files or components]
Do NOT: [things to explicitly avoid]
Definition of done: [unit tests passing / PR ready / deployed to staging]
Priority: [critical path / nice to have]

Effective Aider instructions:
/add src/components/Button.tsx src/styles/button.css tests/button.test.ts

Task: 'Refactor the Button component to use the new design system tokens from src/tokens.ts instead of hardcoded hex values. Update all instances. Ensure all existing tests still pass. Add a test for the new focus-visible state.'

Reclaim setup for deep work: 1. Block 9–11am as 'Focus Time' Habit (defend daily). 2. Import tasks from Asana/Linear with time estimates. 3. Set Meetings window to 2–5pm only. 4. Enable buffer: 10 min after every meeting. 5. Set Fridays as No Meeting day. 6. Add 30min Daily Review habit at 4:30pm. Result: calendar fills with real work automatically, meetings squeeze into 3-hour windows.

Superhuman writing style setup: Spend 5 minutes writing 5 real replies you've recently sent — paste them as style examples in Superhuman's AI settings. The AI learns: your greeting style, signature phrases, how formal/casual you are, your typical email length, and how you structure action items. The better you train it, the less editing each AI draft needs.

Beehiiv newsletter issue brief: Issue number: [N], Subject line options: [Option A — curiosity-based | Option B — direct benefit], Preview text: [expands on subject, 40 chars], Opening hook (50 words): [specific situation/data point your readers recognise], Main section [300 words]: [insight + example + specific takeaway], Practical tip section: [5 bullets, each under 25 words], CTA: [one specific action: reply/share/upgrade/click], Closing: [personal note in your voice].

Hemingway editing pass: 1. Paste your draft. 2. Fix all red sentences first (break into two sentences). 3. Simplify yellow sentences (remove one clause). 4. Replace adverbs with stronger verbs (blue). 5. Convert passive to active voice (green). 6. Check overall grade level — target your audience's level. 7. Readability score 60–70 = strong content. 8. Run a second pass for any regressions from edits.

Semrush content brief workflow: 1. Keyword Magic: find 'main keyword' + 3–5 related LSI keywords. 2. SERP Analysis: study top 3 results for content length, heading structure, FAQ sections. 3. ContentShake: generate brief based on live SERP data. 4. On-Page Checker: run existing content for recommendations. 5. Monitor: add new content to Position Tracking after publishing. 6. Review progress in 4–6 weeks.

Ahrefs content gap analysis workflow: 1. Site Explorer → Competitor domain → Organic Keywords. 2. Content Gap: enter 3 competitors → find keywords all rank for but you don't. 3. Filter: Position 1–10, Volume 100+. 4. Cluster by topic using Parent Topic. 5. Prioritise by traffic potential × your site's DR vs ranking page's DR. 6. Build content calendar from prioritised gaps.

Speechify listening protocol for professionals: 1. First pass at 2× speed — get the structure and main ideas. 2. Second pass at 1.5× on key sections — deepen understanding. 3. Take notes only on second pass — not during first listen. 4. Use AI summary before first pass for long documents — skip if not worth full listen. Target: process 50 pages of text in 30 minutes vs 2 hours traditional reading.

VEED production checklist: 1. Upload raw video. 2. Audio: Clean Audio → remove noise. 3. Subtitles: Auto Subtitle → correct errors → style with brand font/colour. 4. Eye Contact AI: ON if talking-head. 5. Trim: remove silence and dead air. 6. Export: 1080p MP4 for social. 7. Optional: Translate subtitles for regional distribution. Total time: 30 minutes for polished professional video.

Adobe Express design brief: Output: [Instagram post 1080×1080 / YouTube thumbnail 1280×720 / LinkedIn banner], Brand: [company name, primary colour #HEX, secondary colour #HEX, font style], Content: [headline text, supporting text, CTA button text], Key image: [product photo / headshot / icon], Style: [minimal / bold / elegant / playful], Feel: [professional / energetic / calm / urgent].

Fliki blog-to-video brief: Blog URL or script: [paste content], Target platform: [YouTube/Reels/LinkedIn], Duration target: [60s/3min/5min], Voice: [specific language and accent], Tone: [educational/inspiring/entertaining], Brand colours: [HEX], Logo: [upload file], Music style: [upbeat/calm/corporate], CTA at end: [subscribe/visit website/follow].

Linear issue template: Title: [verb + object + context, e.g. 'Fix: Payment fails when INR amount exceeds 1L']
Description: Background: [why this matters]. Steps to reproduce: 1. [step]. 2. [step]. Expected: [correct behaviour]. Actual: [current wrong behaviour]. Acceptance criteria: [ ] [measurable success condition 1]. [ ] [measurable success condition 2]. Priority: [Urgent/High/Medium/Low]. Estimate: [points/hours].

ProWritingAid editing pass for business content: 1. Grammar pass (fix errors). 2. Style report (address overused words — target < 5 instances). 3. Sticky sentences (replace at least half). 4. Readability score (target 60–70 Flesch). 5. Clichés (eliminate or replace with specific examples). 6. Consistency check (especially for proper nouns and technical terms). Time: 30 minutes for a 2,000-word document.

Clearscope content optimisation workflow: 1. Run report for target keyword. 2. Check A+ grade requirements (term list). 3. Outline headings from top-5 competitor structure. 4. Write draft in Google Docs with Clearscope open. 5. Include all Grade A terms naturally (not forced). 6. Reach minimum word count recommendation. 7. Get to A grade before publishing. 8. Track ranking after 60 days.

Tome pitch deck brief: Company name: [name], What you do: [one sentence], Problem: [specific pain point with data], Solution: [how you solve it], Market size: [TAM/SAM/SOM], Traction: [key metrics], Business model: [how you make money], Team: [key backgrounds], Ask: [amount seeking and use of funds], Style: [investor/corporate/educational/storytelling], Audience: [VC/enterprise buyer/board/general public].

Podcastle episode production workflow: 1. Setup: ask guest to use headphones + cardioid mic. 2. Record: local recording ON (each track separate). 3. Post-recording: AI Noise Removal → Silence Removal → Auto Level. 4. Edit: cut filler words from transcript view. 5. Add music: intro/outro 10s, underlay 15% volume. 6. Export: WAV for hosting, MP3 for distribution. 7. Transcription: export for show notes, social clips, blog.

HyperWrite usage pattern for maximum value: 1. Install browser extension. 2. Open Gmail or Google Docs. 3. Start typing your email/doc naturally. 4. When HyperWrite suggests completion — press Tab if it captures your intent, keep typing if not. 5. Use the highlight menu (select text → right click) for Rewrite/Simplify. 6. Use AI Agent for any research question in the same tab. Zero context switching.

Audio enhancement workflow: 1. Record in your normal home environment (don't wait for 'perfect conditions'). 2. Upload raw recording to Adobe Podcast Speech Enhancement. 3. Download enhanced audio. 4. Compare waveforms — should show reduced noise floor. 5. Use in your video/podcast editor. Total added time: 2 minutes per recording. This workflow makes decent home recording sound professional.

Creatify video ad brief: Product: [name and one-line description], USP: [primary benefit in 8 words], Target audience: [specific demographic], Hook (first 3s): [problem statement / shocking stat / question / demonstration], Ad structure: [Hook → Problem → Solution → Social proof → CTA], CTA: [Shop now / Try free / Learn more / Get discount], Platform: [TikTok/Instagram/YouTube], Style: [UGC casual / professional presenter / product demo].