Stop prompting from scratch.Deploy prompts that already work.

Product: [PRODUCT NAME/DESCRIPTION]
Style: photorealistic studio product photography, softbox lighting from top-left, subtle rim light, shallow depth of field (f/2.8 look)
Surface: [matte concrete / brushed metal / warm wood]
Background: gradient [color 1] to [color 2], seamless, no visible horizon line
Camera: 85mm lens, eye-level, centered composition, slight 15-degree angle
Details to emphasize: [texture, logo, key feature]
Mood: premium, minimal, high-contrast shadow
Negative prompt: no text, no watermark, no clutter, no extra props
Aspect ratio: [1:1 / 4:5 / 16:9] --v 6 --style raw

Design a background texture for a social media carousel slide.

Brand colors: [HEX 1], [HEX 2], [HEX 3]
Style: abstract, soft gradient mesh with subtle geometric line accents, low visual noise so text overlays remain readable
Composition: 60% empty space in the center-left for text overlay, visual interest pushed to the edges
Texture: fine grain, no harsh gradients, no banding
Format: vertical 4:5, flat design, no 3D rendering, no photography
Consistency note: this exact style and color set will be reused across a 10-slide series --seed [FIXED SEED NUMBER]

Generate 3 faceless YouTube thumbnail concepts for a video titled: "[VIDEO TITLE]"

For each concept, specify:
- Focal object or icon (no people, no faces, no text baked into the image)
- Color contrast pair (background vs. focal object) for maximum thumbnail pop at small size
- Composition (rule of thirds placement)
- Emotional trigger the visual should create (curiosity / urgency / shock / relief)
- One-line rationale for why this concept earns a click

Constraint: every concept must work with bold text overlay added afterward in an editor, so leave 25% negative space at the top or bottom.

Style: isometric vector illustration, 3/4 perspective, flat shading with soft ambient occlusion
Subject: [e.g., a server rack connected to a cloud icon by glowing data lines]
Color palette: deep navy (#0F172A) base, electric blue (#2563EB) accents, off-white (#F8FAFC) highlights
Line work: clean 2px outlines, rounded corners on all objects
Background: solid navy or transparent
Details: small floating UI cards or icons to suggest a dashboard/automation theme
Do not include: text, human figures, realistic textures, gradients heavier than a subtle two-tone shade
Aspect ratio: 16:9, blog header format

Create a cinematic concept art prompt for an original character (no existing copyrighted character).

Character: [role/archetype, e.g., "a rogue data courier in a rain-soaked megacity"]
Wardrobe: [key garments and materials]
Pose: dynamic three-quarter stance, mid-motion
Lighting: cinematic rim light, teal and orange color grade, volumetric fog
Camera: low angle, 35mm lens look, shallow depth of field
Environment: [brief setting description]
Mood: [tense / heroic / melancholic]
Render style: hyper-detailed digital painting, concept-art quality, not photorealistic
Negative prompt: no text, no logos, no watermark, no extra limbs

Act as an SEO content strategist. Target keyword: "[KEYWORD]"

1. Identify the likely search intent (informational / commercial / navigational / transactional) and justify it in one sentence.
2. Propose 3 title options under 60 characters, each with the keyword near the front.
3. Build an H2/H3 outline (6-8 sections) that fully answers the intent, including one "common mistakes" section and one FAQ section with 4 questions.
4. List 8 secondary/semantic keywords to naturally weave in.
5. Suggest a meta description under 155 characters.

Write a [30-45 second] faceless short-form video script about: "[TOPIC]"

Structure:
1. Hook (0-3 sec): a pattern-interrupt line or bold claim, under 12 words.
2. Setup (3-8 sec): why this matters to the viewer specifically.
3. Payoff (8-30 sec): 3 concrete points, each one sentence, no fluff.
4. CTA (final 3-5 sec): one clear action (follow / comment / save).

Also provide:
- A matching on-screen text overlay for each beat (short, punchy)
- A suggested pixabay_search_term (4-8 lowercase words, ending in "no people") for stock B-roll
- Tone: [energetic / calm-authoritative / conversational]

Source content (paste below): a blog post / video transcript / podcast summary.

Repurpose it into 10 distinct assets:
1. A 3-tweet thread (hook + 2 value tweets)
2. A LinkedIn post (under 150 words, personal framing)
3. A carousel outline (5 slides, one idea per slide)
4. Three short-form video hook lines (different angles)
5. An email newsletter blurb (60 words + CTA)
6. A Pinterest/Instagram caption with 5 relevant hashtags
7. A one-sentence "quote card" pull-quote

For each asset, label the format and keep the core insight intact — do not just shorten the original, reframe it for that platform's native behavior.

Source content:
"""
[PASTE SOURCE CONTENT HERE]
"""

Product/service: "[NAME + ONE-LINE DESCRIPTION]"
Target audience: [WHO]
Primary pain point: [PAIN]
Primary outcome/promise: [OUTCOME]

Write landing page copy in this structure:
1. Headline (under 10 words, outcome-led)
2. Sub-headline (1 sentence, names the audience + mechanism)
3. 3 bullet-point benefits (not features — outcomes)
4. "How it works" in 3 numbered steps
5. One objection-handling paragraph (address the #1 reason people don't buy)
6. Closing CTA line + button text (2-4 words)

Tone: [direct and confident / friendly and reassuring]. No hype words like "revolutionary" or "game-changing."

Niche: "[YOUR NICHE]"
Platform(s): [Instagram / Facebook / YouTube Shorts]
Posting cadence: [e.g., 5 posts/week]

Build a 30-day content calendar as a table with columns:
Day | Content Pillar (pick from: Educate / Entertain / Inspire / Sell / Community) | Post Idea (one line) | Format (Reel / Carousel / Static / Story)

Rules:
- Rotate pillars so no single pillar exceeds 40% of the month.
- Include exactly 4 "Sell" posts across the month, spaced at least 5 days apart.
- Group ideas so every 7 days forms a loosely connected mini-theme.

Act as a senior Python engineer doing a code review. I will paste a function below.

1. Refactor it for readability and PEP 8 compliance without changing its external behavior or function signature.
2. Add a full docstring (Google style): description, Args, Returns, Raises.
3. Add inline type hints if missing.
4. List any edge cases the current implementation does NOT handle, without fixing them unless I ask.
5. Do not add new dependencies.

Function:
"""
[PASTE FUNCTION HERE]
"""

I'm debugging an issue. Here is what I know:

Expected behavior: [WHAT SHOULD HAPPEN]
Actual behavior: [WHAT ACTUALLY HAPPENS]
Error message / stack trace: [PASTE IF ANY]
Relevant code: [PASTE CODE]
Environment: [language/framework/OS version]

Do the following, in order:
1. List the 3 most likely root causes, ranked by probability, with one-sentence reasoning each.
2. For the top cause, propose the smallest possible test I can run to confirm or rule it out.
3. Only after I confirm the cause, propose a fix. Do not propose a fix before that.

Framework: [FastAPI / Express / Flask / Django REST]
Endpoint: [METHOD] [PATH], purpose: [WHAT IT DOES]
Request body fields: [FIELD: TYPE, ...]
Response shape (success): [DESCRIBE FIELDS]

Generate:
1. The route/handler code with input validation (reject malformed input with a clear 4xx error).
2. Structured error responses for: missing fields, invalid types, and one relevant business-logic failure case.
3. A short comment block above the handler explaining the auth requirement (if any).
4. One example curl request and expected JSON response.

Do not include database connection boilerplate — assume a repository/service layer already exists.

I'll paste a script that runs on a schedule (cron / Azure Automation / GitHub Actions). Please:

1. Summarize what it does in plain English, step by step, in the order it executes.
2. Identify every external dependency (API, file path, environment variable, credential) it relies on.
3. Flag any spot where a failure would fail silently (no error surfaced, no log, no alert) rather than crashing loudly.
4. Suggest minimal logging or try/except additions to surface those silent failures — do not restructure the whole script.

Script:
"""
[PASTE SCRIPT HERE]
"""

Database: [PostgreSQL / MySQL / SQL Server]
Here is a query that's running slowly:
"""
[PASTE SQL QUERY HERE]
"""
Table sizes (approx): [TABLE: ROW COUNT, ...]
Existing indexes (if known): [LIST OR "unknown"]

1. Explain what the query is doing in plain English.
2. Identify the most likely performance bottleneck (e.g., missing index, N+1 pattern, unnecessary full scan, non-sargable WHERE clause).
3. Rewrite the query for better performance without changing its output.
4. Suggest one index (with exact CREATE INDEX statement) if that would help, and explain the tradeoff of adding it.