Explore all AI prompts tagged with #AI Prompt.
Midjourney v6portrait of a man in a red santa hat, snow falling, golden hour rim light, 85mm f/1.8, shallow depth of field, warm color grade, photorealistic
A warm winter portrait recipe that leans on golden-hour rim light and an 85mm depth of field to separate the subject from falling snow. Works at 4:3 and 1:1. Swap the hat colour or subject for fast variations without losing the light.
SDXLcozy log cabin interior at christmas, fireplace glow, decorated tree with warm fairy lights, wide angle 24mm, soft window light, photorealistic, high detail
An interior scene built around two believable light sources — fireplace and window — so the fairy lights read as accents instead of noise. Great for greeting cards and seasonal banners; keeps detail without the HDR-everything look.
Midjourney v6candid photo of a couple walking under one umbrella in light rain, wet street reflections, 50mm, evening city bokeh, muted teal and amber grade
A two-subject composition that solves the spacing problem: the umbrella forces natural closeness while street bokeh keeps focus on the pair. The muted teal-amber grade is specified so a whole series stays colour-consistent.
Midjourney v6anime key visual, character on a rooftop overlooking a neon city at night, wind in hair, painterly clouds, dramatic rim light, makoto shinkai style lighting
A key-visual composition with a named lighting reference so the style lands deliberately. The rooftop framing gives you a foreground anchor and an excuse for dramatic rim light; swap the skyline for fantasy variants.
SDXLskincare bottle floating above rippling water surface, soft studio key light, pastel seamless background, water droplets frozen mid-air, commercial product photography, 100mm macro
A packshot recipe with a frozen-motion water element that reads premium without props. The 100mm macro spec keeps label text sharp — the detail that usually breaks generated product shots. Works for any bottle-shaped product.
Fluxalpine valley filled with low fog at dawn, first light hitting peaks, layered ridgelines, 70-200mm compression, muted blue-gold palette, ultra detailed
A landscape prompt built on telephoto compression and layered fog rather than saturation. The blue-gold palette constraint stops the sunrise from going radioactive, and ridgeline layering gives depth that flat wide shots miss.
Midjourney v6professional headshot of a woman in a navy blazer, large window light from camera left, clean grey seamless, 85mm f/2, natural retouch, confident slight smile
A LinkedIn-grade headshot recipe: one big soft source, one seamless, one lens. Specifying 'natural retouch' keeps skin texture believable. Batch-generate a team's worth by swapping subject and blazer colour only.
Fluxstreet style editorial photo of a man in an oversized wool coat, crossing a city street, motion blur in background traffic, overcast soft light, 35mm, fujifilm color science
An editorial look that uses overcast light as a giant softbox and background motion blur for energy. The film-stock reference locks the grade. Works as a series — change the outfit, keep the recipe.
Runwayslow dolly-in toward a rain-streaked window at night, city lights defocused beyond the glass, anamorphic lens flares, moody blue grade, 24fps, shallow focus
A five-second mood piece that works because the camera move is slow and singular. Anamorphic flares and defocused city light give production value with zero subjects to deform. Ideal opener for a moody product or title reveal.
Runway[white sneaker] rotating 360 degrees on a glossy black turntable, dramatic top key light, dust particles in light beam, seamless loop, 4k, product commercial style
A loopable turntable shot with one key light and floating dust for atmosphere. The seamless-loop instruction matters: it makes the clip usable as a background hero on landing pages without a visible cut.
Klingaerial drone shot sweeping along a turquoise coastline with white cliffs, morning light, gentle forward motion, waves rolling in slow motion, cinematic color grade
A travel-reel staple: forward drone motion, morning light, slow-motion surf. The gentle-motion constraint keeps the AI from producing nausea-cam. Pairs well with a mid-reel cut to street-level footage.
isometric 3d animation of a saas dashboard assembling itself, cards sliding into place, soft shadows, calm easing, clean white background, subtle camera drift
An explainer bed with calm easing and a white background so UI text stays legible under voiceover. The assemble-itself motif communicates 'product coming together' without needing a single real screenshot.
Midjourney v6minimal logo, letter D monogram formed by circuit traces, single color, flat vector, strong negative space, works at 16px favicon size, no gradients
A developer-tool mark built from a single-letter circuit motif. The 16px constraint in the prompt is the trick — it forces simplification that survives favicon duty. Ask for three weights to test scalability.
Midjourney v6vintage hand-drawn crest logo for a french bistro, crossed cutlery and wheat sprigs, circular badge, letterpress texture, two-color, 1920s engraving style
A badge-style restaurant mark with era and technique pinned to the 1920s engraving look. Two-colour constraint keeps print costs real. The crossed-cutlery motif reads 'bistro' from across the street.
Audit this React component for unnecessary re-renders. List each cause (prop identity, context, state colocation), rank by measured impact, then propose the minimal fix for each — memo, callback stability, or state restructuring. Show before/after code. Do not suggest rewrites larger than the problem. [PASTE COMPONENT]
▸ ready to runAudit this React component for unnecessary re-renders. List each cause (prop identity, context, state colocation), rank by measured impact, then propose the minimal fix for each — memo, callback stability, or state restructuring. Show before/after code. Do not suggest rewrites larger than the problem. [PASTE COMPONENT]
A performance-review prompt that forces ranking by impact before any fix, which stops the usual memo-everything reflex. The minimal-fix constraint keeps diffs reviewable. Paste one component at a time for best results.
Review every useEffect in this file. For each: state what it synchronises, whether it should be an event handler instead, and whether the dependency array is correct. Flag effects that should not exist. Output a table: effect | purpose | verdict | fix. [PASTE FILE]
▸ ready to runReview every useEffect in this file. For each: state what it synchronises, whether it should be an event handler instead, and whether the dependency array is correct. Flag effects that should not exist. Output a table: effect | purpose | verdict | fix. [PASTE FILE]
Turns the vaguest part of React — effects — into a table audit with a verdict per effect. The 'should this be an event handler' question catches the majority of real-world effect misuse. Great pre-review pass.
Write a pandas pipeline that ingests this CSV description, handles missing values with per-column strategies you justify, normalises dtypes, and outputs a validated DataFrame. Use method chaining, type hints, and add three assertion checks that would catch silent data corruption. [DESCRIBE COLUMNS]
▸ ready to runWrite a pandas pipeline that ingests this CSV description, handles missing values with per-column strategies you justify, normalises dtypes, and outputs a validated DataFrame. Use method chaining, type hints, and add three assertion checks that would catch silent data corruption. [DESCRIBE COLUMNS]
A data-cleaning prompt that demands per-column justification and built-in assertions — the difference between a script and a pipeline you can trust next month. Method chaining keeps the result readable in review.
Here is a slow PostgreSQL query and its EXPLAIN ANALYZE output. Walk the plan node by node, identify the dominant cost, and propose fixes in order of least invasive: query rewrite, then index, then schema change. Estimate the impact of each. Never suggest an index without naming its write cost. [QUERY + PLAN]
▸ ready to runHere is a slow PostgreSQL query and its EXPLAIN ANALYZE output. Walk the plan node by node, identify the dominant cost, and propose fixes in order of least invasive: query rewrite, then index, then schema change. Estimate the impact of each. Never suggest an index without naming its write cost. [QUERY + PLAN]
A query-tuning prompt ordered by invasiveness so you try rewrites before adding indexes. Requiring the write-cost caveat on every index suggestion filters out advice that would just move the pain to inserts.
Given this stack trace and code context, produce a ranked list of at most four hypotheses for the root cause. For each: the evidence for, the evidence against, and the single cheapest experiment that would confirm or kill it. Do not propose any fix until one hypothesis survives. [TRACE + CONTEXT]
▸ ready to runGiven this stack trace and code context, produce a ranked list of at most four hypotheses for the root cause. For each: the evidence for, the evidence against, and the single cheapest experiment that would confirm or kill it. Do not propose any fix until one hypothesis survives. [TRACE + CONTEXT]
A debugging discipline in prompt form: hypotheses with evidence both ways and a cheapest-experiment step, with fixes explicitly forbidden until confirmation. Cuts wild-goose-chase time dramatically on unfamiliar codebases.
This function is too long and untested. Produce: (1) a characterisation test suite that pins current behaviour including its bugs, (2) a seam-by-seam extraction plan, (3) the refactored version. Behaviour changes are forbidden — flag suspected bugs in comments instead of fixing them. [PASTE FUNCTION]
▸ ready to runThis function is too long and untested. Produce: (1) a characterisation test suite that pins current behaviour including its bugs, (2) a seam-by-seam extraction plan, (3) the refactored version. Behaviour changes are forbidden — flag suspected bugs in comments instead of fixing them. [PASTE FUNCTION]
A refactoring prompt that puts characterisation tests before any code movement and outlaws behaviour changes — suspected bugs get flagged, not silently fixed. The output is a safe, reviewable migration rather than a rewrite gamble.
For this function, identify the algebraic properties it should satisfy (idempotence, inverse pairs, invariants, ordering). Write property-based tests for each using the idiomatic library for the language. Include two deliberately nasty generators: adversarial strings and boundary numerics. [PASTE FUNCTION]
▸ ready to runFor this function, identify the algebraic properties it should satisfy (idempotence, inverse pairs, invariants, ordering). Write property-based tests for each using the idiomatic library for the language. Include two deliberately nasty generators: adversarial strings and boundary numerics. [PASTE FUNCTION]
Moves testing from example-listing to property thinking: the prompt asks which algebraic laws the code must obey, then encodes them with hostile generators. Finds the edge cases table-driven tests always miss.
Review this schema migration for a production Postgres database under load. Check: lock levels per statement, table rewrite triggers, index build strategy, and rollback path. Output a go/no-go verdict with the exact statement that carries the most risk and a safer sequencing if one exists. [PASTE MIGRATION]
▸ ready to runReview this schema migration for a production Postgres database under load. Check: lock levels per statement, table rewrite triggers, index build strategy, and rollback path. Output a go/no-go verdict with the exact statement that carries the most risk and a safer sequencing if one exists. [PASTE MIGRATION]
A pre-deploy gate for schema changes: every statement gets a lock-level check and the riskiest one is named explicitly. The safer-sequencing requirement regularly converts a table-locking migration into a zero-downtime one.
Write a 4-email launch sequence for [PRODUCT] aimed at [AUDIENCE]. Email 1: problem agitation, no product mention. Email 2: mechanism reveal. Email 3: social proof + offer. Email 4: deadline close. One CTA per email, subject lines under 45 characters, plain-spoken tone, no exclamation marks.
▸ ready to runWrite a 4-email launch sequence for [PRODUCT] aimed at [AUDIENCE]. Email 1: problem agitation, no product mention. Email 2: mechanism reveal. Email 3: social proof + offer. Email 4: deadline close. One CTA per email, subject lines under 45 characters, plain-spoken tone, no exclamation marks.
A launch skeleton with a fixed narrative arc — agitate, reveal, prove, close — and hygiene rules that keep the copy out of the promotions-tab voice. Fill the brackets and you have a shippable draft to edit, not a blank page.
Generate 12 scroll-stopping hooks for a cold Facebook audience for [OFFER]. Mix formats: 3 contrarian claims, 3 specific-number results, 3 problem call-outs, 3 curiosity gaps. Max 12 words each. Then pick the 2 you would test first and say why.
▸ ready to runGenerate 12 scroll-stopping hooks for a cold Facebook audience for [OFFER]. Mix formats: 3 contrarian claims, 3 specific-number results, 3 problem call-outs, 3 curiosity gaps. Max 12 words each. Then pick the 2 you would test first and say why.
A hook-generation prompt with enforced format variety so you get a real test matrix instead of twelve rephrasings of one idea. The self-selection step at the end doubles as a quality filter you can argue with.
