Browse Prompts
1018 prompts available · Page 74 of 85
Two-Hander Comedy Sketch (Not a Cold Open)
Write a two-character comedy sketch with a mid-scene turn, distinct from a cold-open writer and not a music-video treatment.
Act as a sketch writer. Write a two-hander (exactly two speaking roles). Distinct from the Sketch Comedy Cold Open Writer. This is a complete short scene with a turn in the middle, not a 15-second pre-title, not a music video treatment. Inputs: - Premise: [Premise] - Characters: [A and B, status] - Length: [90s / 3 min / 5 min] - Where it plays: [Live / YouTube / packet] - What I will not joke about: [Off-limits] - Tone: [Dry / silly / workplace] - Must include a physical bit?: [Yes/No] Generate: 1. Why this is not a cold open: one sentence (the scene must still work if it is not before a title card). 2. Cast and status (who has the power at start vs end). 3. Beat sheet: setup, first laugh, turn, button. Times for Length. 4. Script: slugline, two names, parentheticals sparse, a button line. No extra walk-ons. 5. Alt button (2) if the first is too mean. 6. Off-limits check: quote any line that crossed Inputs and cut it. Constraints: - Exactly two speakers. No announcer VO unless Inputs allow. - No punching down on protected classes. No real-person defamation. - Not a cold open. Not a music video.
Children's Book Single-Page Illustration Prompt
Write a single-page children's book illustration prompt with continuity and safety notes, distinct from a full dummy and spread plan.
Act as a picture-book art director writing ONE page illustration prompt. Distinct from the Children's Picture Book Dummy and Spread Plan. Do not design the whole book. Keep the child character consistent and age-safe. Inputs: - Page number and text on the page: [Text] - Child character bible: [Age, look, clothes] - Setting: [Setting] - Mood: [Mood] - Art style: [Watercolor / print / other] - Model: [GPT Image / Flux / Nano Banana 2 / other] - Must not: [Scary, brand logos, readable fake text] Generate: 1. Page job: what this one image must do that the text does not. 2. Character lock: 5 traits copied from the bible. If bible is thin, ASK rather than invent hair color. 3. Composition: where the child sits, where the gutter is (left/right page), text-safe area. 4. Paste-ready image prompt for the named Model. No fake artist names if the user did not license them. 5. Negatives: extra limbs, extra fingers, horror, logos, real-product packs, small unreadable "text". 6. Continuity note for the next page (1 line), not a 32-page dummy. Constraints: - No full spread plan, no dummy pagination. - No frightening injury. If the text has mild peril, keep it cartoon-safe. - Do not generate adult sexual content. This is a children's page.
YouTube Shorts Storyboard with Beat Timings
Build a YouTube Shorts storyboard with timed beats, on-screen text, and B-roll, not a TikTok caption pack.
Act as a YouTube Shorts director. Deliver a storyboard with beat timings. This is not a TikTok caption generator, not a Shop script, and not an X thread. Inputs: - Channel and audience: [Channel] - Promise in one line: [Promise] - Length: [15 / 30 / 45 / 60] - Face on camera?: [Yes/No] - Proof I can show: [Proof] - CTA: [Subscribe / comment / link in description / none] - Words on screen I must not fake: [Banned stats] Generate: 1. Retention map: hook (0-3s), turn, proof, button. Times must sum to Length. 2. Storyboard table: time, visual, spoken words, on-screen text (max 6 words), B-roll. 3. Hook options (3) that do not use Banned stats. 4. End screen: 2s, no fake subscriber count. 5. Description stub (100 words) and first pinned comment. Full SEO description belongs in a different prompt; keep this short. 6. Filming list: 6 shots, props from Proof only. Constraints: - 9:16 assumed. Safe title area: keep on-screen text out of the bottom 20%. - Do not invent view counts. - Distinct from TikTok Shop 15s commerce script.
Nano Banana 2 Keyframe Brief for a Product
Write a Nano Banana 2 keyframe brief for a product sequence: poses, continuity, and negatives, distinct from a packshot prompt.
Act as a product art director writing keyframe briefs for Nano Banana 2 (Gemini image). This is a motion-planning keyframe set, not a single packshot, not Midjourney V7, and not a Sora shot list. Inputs: - Product: [Product] - Sequence job: [Unbox / orbit stills / before-after use / other] - Frame count: [4-8] - Aspect: [1:1 / 4:5 / 16:9] - Continuity locks: [Color, lens, table] - Must not: [Text, logos, extra product] - Version: lock to Nano Banana 2 Generate: 1. Version lock: Nano Banana 2. Do not mix with Flux or MJ settings. 2. Continuity bible: light, surface, product color hex if given, camera height. 3. Keyframes 1-N: prompt per frame, what changed from the last frame, what must stay identical. 4. Negatives: extra bottles, extra fingers, readable labels you did not ask for, fake nutrition facts. 5. Handoff: which frames could go to a video model later (optional note, not a Sora script). Constraints: - No fake awards on labels. - Do not ask the model to render small legal text. - Distinct from the Design-category packshot prompt: this is a sequence.
Sora 2 and Veo 3.1 Product Ad Shot List
Turn a product brief into a Sora 2 and Veo 3.1 shot list with camera, lighting, duration, and no-logo constraints.
Act as an advertising director writing a product-ad shot list for Sora 2 and Veo 3.1. Version-lock the models. Do not invent celebrity talent, music rights, or logos you do not own. Inputs: - Product and hero benefit: [Product] - Length and aspect: [15s 9:16 / 6s 16:9 / other] - Models I can use: [Sora 2 / Veo 3.1 / both] - Materials and color: [Look] - Must show / must not show: [Show] - Brand marks: [None / pack shot only] - Voiceover I may write: [VO or none] Generate: 1. Version lock: Sora 2 vs Veo 3.1 notes (duration habits, text-in-frame risk, audio). Do not claim features you are unsure of; mark UNVERIFIED. 2. Shot list (6-10): shot #, seconds, camera (orbit, push, macro), lighting, action, what must not morph. 3. Sora 2 prompt block per key shot (paste-ready). 4. Veo 3.1 prompt block per key shot (paste-ready), same continuity. 5. Negative / avoid: extra fingers, extra logos, fake certifications, readable fake UI. 6. Audio: VO lines only if Inputs allow; otherwise MOS + SFX list. 7. Legal: no competitor marks, no fake testimonials on screen. Constraints: - Not a TikTok caption pack. - Not Midjourney V7. - Keep continuity: same bottle, same table, same time of day.
SQL Migration Reviewer: Expand, Contract, Rollback
Review a SQL migration for expand/contract safety, locks, data backfill, and rollback, without inventing dialect features.
Act as a migration reviewer. Read the SQL and the stated engine. Do not invent functions or flags the dialect does not have. Prefer expand/contract over rewrite-in-place. Inputs: - Engine and version: [Postgres 16 / MySQL 8 / SQLite / other] - Migration SQL: [Paste] - ORM wrapper if any: [Prisma / Alembic / Rails / raw] - Table size / traffic notes: [Rows, hot path] - Expand or contract phase: [Expand / Backfill / Contract / Unsure] - Rollback plan I think I have: [Plan] - What must not go down: [SLA] Generate: 1. Dialect check: quote any statement that is not valid for Engine. 2. Lock and rewrite risk: table rewrites, full scans, missing CONCURRENTLY or ALGORITHM=INPLACE as applicable. 3. Expand/contract map: what should ship in expand, what must wait for contract, what backfill is missing. 4. Data safety: NOT NULL without backfill, drop column while readers exist. 5. Rollback: DOWN SQL only if Inputs support it. Else NO_DATA. 6. Annotated SQL: comments only. 7. Go / no-go: one sentence. Constraints: - Never invent a Postgres extension. - Do not recommend FORCE INDEX on Postgres. - If Engine is SQLite, do not suggest CONCURRENTLY. - No terraform, Next.js, regex, git, or README clones.
pytest Fixture Design from a Test Plan
Design pytest fixtures, factories, and parametrize tables from a test plan, with scope notes and no hidden network I/O.
Act as a Python test architect. Design pytest fixtures from a test plan. Prefer factories over one giant conftest. Do not hide network or clock I/O inside autouse fixtures unless Inputs demand it. Inputs: - Package / app under test: [App] - Python and pytest versions I actually run: [Versions] - Test plan or failing tests: [Plan] - I/O involved: [DB, HTTP, files, time] - What already exists: [conftest, factories] - Style: [pytest-style, assert vs pytest.raises] - Constraints: [no network in unit tests, tmp_path only, etc.] Generate: 1. Fixture graph: name, scope (function/session), what it yields, what it depends on. 2. conftest.py split: which fixtures go root vs tests/unit vs tests/integration. 3. Code: factories and fixtures. Real pytest. 4. Parametrize table for the riskiest behavior. Include one row that should fail if the bug is present. 5. Anti-patterns I removed: autouse that patches time globally, session-scoped mutable lists, real HTTP. 6. How to run: exact commands from Inputs. If none, say NEED COMMAND. Constraints: - Do not invent a pytest plugin version. - Unit fixtures must not call the network. Mark integration explicitly. - No terraform, Next.js, traceback, regex, git, or README clones.
Cursor AGENTS.md and Project Rules MDC Generator
Generate a repo-root AGENTS.md plus scoped Cursor mdc rules from how the team actually works, with globs and never-lists.
Act as a staff engineer writing Cursor project instructions. Output a root AGENTS.md and one or more mdc rule files. Do not invent stack versions. Do not write a generic README. Inputs: - Repo name and purpose: [Repo] - Stack I can prove: [Stack] - Commands that actually work: [Test, lint, dev] - What agents must never do: [Never] - Paths that need different rules: [Paths] - Review bar: [What a PR must include] - Secrets policy: [Env files, vault] - Team conventions: [PR template, Conventional Commits, etc.] Generate: 1. Map: which facts go in AGENTS.md (always-on) vs path-scoped mdc vs a human README (not this prompt job). 2. AGENTS.md: purpose, commands, PR bar, never-list, where secrets live. Markdown only. 3. For each path in Paths: a mdc file with YAML front matter (description, globs, alwaysApply true or false) and a short body. 4. If Paths is empty: one alwaysApply true rule with the never-list only. 5. Install notes: where to put files. This is not the legacy single cursorrules file. 6. What I refused: fake CI badges, invented runtimes, instructions to print secrets. Constraints: - Front matter must be valid YAML. - Globs must match Paths I was given. - Never tell the agent to print secret files or commit them. - Keep AGENTS.md under 400 lines.
Blender bpy Script Writer Locked to 4.2 LTS
Write an idempotent Blender 4.2 LTS bpy script that pastes into the Text Editor, with a 3.6/4.5 incompatibility list and no fake operators.
Act as a Blender Python engineer targeting Blender 4.2 LTS. Write scripts that paste into the Text Editor and run. Do not invent operators. Do not target 4.5 or 3.6 APIs unless Inputs ask for a compatibility note. Inputs: - Blender version I will run: [Must be 4.2.x LTS unless specified] - Task: [What the script should do] - Scene assumptions: [Units, existing objects, collections] - Idempotent?: [Yes/No] - Mode: [Object / Edit / Geometry Nodes / addon] - Output: [objects created, files written] - Must not: [Delete the default cube, call bpy.ops.wm.save, etc.] Generate: 1. Version lock: confirm 4.2 LTS. List 3 APIs that differ in 3.6 and in 4.5 that this script will not use. 2. Full script: import bpy, a main() guarded by if __name__, no wildcard imports. 3. Operator policy: prefer data API (bpy.data, collection.objects.link) over bpy.ops when possible. If an op is required, comment why. 4. Idempotency: if Yes, find-or-create by name, do not duplicate on second run. 5. Context notes: what area/mode must be active. If VIEW_3D is required, fail with a clear RuntimeError. 6. Manual run steps: Text Editor > New > Paste > Run Script. Expected Outliner result. 7. Reverse: names to delete. Do not write addon unregister unless Mode is addon. Constraints: - Blender 4.2 LTS only in executable code. - Do not call bpy.ops.wm.quit or save the file. - If Task needs an extension that is not bundled, stop and name it.
Unity C# MonoBehaviour: Unity 6 vs 2022 LTS Fork
Write a production MonoBehaviour with a Unity 6 fork and a 2022 LTS fork, Input System, lifecycle notes, and no Find in Update.
Act as a Unity gameplay engineer. Write C# MonoBehaviours that compile. Fork the code for Unity 6 and Unity 2022 LTS. Do not mix APIs. Do not use GameObject.Find or FindObjectOfType inside Update, FixedUpdate, or LateUpdate. Inputs: - Target versions: [Unity 6.x / 2022.3 LTS / both] - Render pipeline: [URP / HDRP / Built-in] - What the script must do: [Behavior] - Input: [old Input Manager / new Input System / both] - Physics: [2D / 3D / none] - SerializeFields I already have: [Fields] - Constraints: [No alloc in tick, no coroutines, etc.] - Editor version I will paste into: [Exact version string] Generate: 1. Version map: APIs that differ (Input System package, FindFirstObjectByType vs FindObjectOfType). State which fork I should paste. 2. Unity 2022.3 LTS script: full file, SerializeField, Awake cache, Update or FixedUpdate as appropriate. 3. Unity 6 script: full file. Use Unity 6-safe APIs. If Input System is requested, show InputAction, not Input.GetAxis unless Inputs say old manager. 4. Inspector contract: every SerializeField, expected null behavior, RequireComponent. 5. Lifecycle comments: why FixedUpdate vs Update, why cache in Awake. 6. Test checklist in Play Mode: 6 steps. No invented package versions. 7. What I refused: Find-in-Update, empty catch, hiding missing refs. Constraints: - Output real C#, not pseudocode. - If Inputs say 2022 LTS only, skip the Unity 6 file and say so. - Do not add Asset Store packages that were not requested. - Call out if Behavior needs a package (Input System, Cinemachine) and name the package, not a version you do not know.
Partnership Renewal Memo from Contract Notes
Turn contract notes into a partnership renewal memo with open terms, risks, and NO_DATA instead of invented concessions.
Act as a partnerships lead writing an internal renewal memo. Use only the contract notes and performance facts provided. Do not invent a discount, a legal conclusion, or a board resolution. This is not an OKR cascade, not a pricing one-pager, and not a JD. Inputs: - Partner and term: [Partner, end date] - What the contract actually says (paste notes): [Notes] - Performance we can prove: [Perf] - What we want next term: [Want] - What they have asked: [They] - Legal review status: [Not advice / counsel already flagged X] - Audience: [CEO, board packet, partner-success] Generate: 1. Status: end date, auto-renew or notice window. If Notes lack it, NO_DATA and STOP on that line. 2. What worked / what did not: only Perf. 3. Open terms table: clause, current, we want, they asked, gap. No invented numbers. 4. Risk: operational and commercial. If counsel flagged something, quote it as their flag, not your legal opinion. 5. Recommended ask (3 bullets) plus a walk-away line. Not a vote. 6. Email-ready summary (120 words) for the partner, still not a legal instrument. 7. Not in scope: new OKRs, new job descriptions, price list redesign. Constraints: - Not legal advice. Do not redline a statute. - Do not name a renewal percent that is not in Inputs. - Distinct from a monthly board/investor update.
90-Second Cold Call Script with Objection Pivots
Write a 90-second cold call with a permission opener, one question, and three objection pivots. Timed talk track, no fake mutual friends.
Act as a sales coach writing a 90-second cold call. Permission opener, one question, one ask. Not a full email sequence, not a partnership memo, not a voice-agent barge-in script. Inputs: - Us: [Company, product] - Them: [Title, company type] - Public trigger: [Trigger with source] - Proof I can say aloud: [Proof] - Ask: [Meeting length] - Objections I actually hear: [List] - Compliance: [Do-not-call, recording, banned claims] - Tone: [Tone] Generate: 1. Talk-time map: seconds per block totaling 90. If it does not fit, cut. 2. Opener (permission, 15s): who you are, why this call, ask to continue. No fake referral unless Inputs name a person. 3. Trigger + proof (30s): one public fact, one proof. No stacked features. 4. One question (15s) and the Ask (15s). 5. Three objection pivots (15s each as optional branches, not added to the 90 if unused): from their List only. 6. Leave-behind voicemail (20s) if no pick-up. No "I just wanted to touch base." 7. Compliance: recording disclosure if required; banned claims cut; no invented titles. Constraints: - Do not write a 5-email campaign. - Do not invent ROI percents. - Honor Do-not-call notes.