promptstudio
promptstudio's Prompts
Runway Gen-3 Camera Move List for a Product Hero (No Logos or Celebs)
Write a Runway Gen-3 camera-move shot list for a product hero. No logos, celebrities, or trademarked UI.
Act as a motion director writing a Runway Gen-3 camera-move list for a product hero. You write camera-move recipes for Runway Gen-3 Alpha Turbo / Gen-3 from a product brief. You describe dolly, orbit, crane, and handheld language the model can take. You do not name celebrities, logos, or trademarked UI. This is not a Kling loop list, not a Pika demo list, and not a generic Runway product loop. You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs. Inputs: - Runway model lock: [Model] - Product (unbranded description, materials, color): [Product] - Hero surface and set (no brand names): [Set] - Camera moves I allow (dolly in, orbit, crane, static): [Moves] - Duration and aspect I can prove: [Format] - What must never appear: [Never] - Seed or reference still notes I may use (or NONE): [Ref] - Words I must not use: [Banned] - How many shots: [Count or 6] - Lighting I can prove: [Light] Generate: 1. Honesty ledger: model, product materials, set, moves, format, light. Forbidden: logos, celebs, UI chrome, competitor product names. 2. Shot table: Count rows. Each has move from Moves, duration from Format, product placement from Product, light from Light. 3. Prompt block per shot: subject, set, camera move, negative (Never). No celebrity, no logo text. 4. What this is not: Kling loop, Pika demo, orbit-only loop. Say camera-move list for a still-to-motion hero. 5. Ref: only Ref. If NONE, do not invent a seed. 6. Negatives: Never plus logos, celebs, readable packaging type unless Product quoted type. 7. Run notes: Model lock, aspect, seconds. No invented feature names. 8. Compliance pass: quote Banned words, celebrity names, logo requests. Cut them. Constraints: - Runway Gen-3 camera-move list for a product hero. Not Kling, not Pika, not a loop pack. - No logos, celebrities, trademarked UI, or readable brand marks. - Only camera moves listed in Moves. - If Ref is NONE, do not invent a seed number. - No emojis.
Playwright e2e Scaffold from a User Journey (No Invented Selectors)
Scaffold Playwright tests from a pasted user journey. No invented selectors, URLs, or auth secrets.
Act as a QA engineer writing a Playwright test scaffold from a user journey. You write Playwright Test TypeScript (v1.47 style) from a journey the user pastes. You only use selectors, roles, and URLs that appear in Inputs. You do not invent data-testid values, email addresses, or storageState files. This is not Cypress and not a unit test. You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs. Inputs: - Playwright version lock: [Version] - Base URL: [Base] - Journey steps (action + selector or role I can prove): [Journey] - Assertions I can prove (text, URL, status): [Assert] - Auth setup I can prove (storageState path or env name): [Auth or NONE] - Projects/browsers I allow: [Browsers] - Fixtures or test ids I actually have: [Fixtures] - What I must never click or call: [Never] - Words I must not use: [Banned] - Timeout overrides I can prove (or NONE): [Timeout] Generate: 1. Honesty ledger: version, base, every selector in Journey, browsers, auth. Forbidden: testids and URLs not listed. 2. playwright.config.ts sketch: testDir, baseURL, projects from Browsers only. Timeout only from Timeout. 3. spec.ts: one test named from the journey. Each Journey step becomes a locator call using the given selector or getByRole name. If a step has no selector, write NOT IN INPUTS and skip that line rather than guessing [data-testid=submit]. 4. Expect: only Assert. Do not add a screenshot assertion unless Inputs asked. 5. Auth: use Auth storageState or env name. If NONE, leave the test unauthenticated and say so. 6. Never: comment a blocklist of Never locators. 7. Run notes: npx playwright test pinned to Version. No invented grep tags. 8. Compliance pass: quote Banned words, invented selectors, Cypress cy.get. Cut them. Constraints: - Playwright Test TypeScript only. Not Cypress, not Selenium, not Playwright Python unless Inputs say so. - Never invent a selector, testid, or URL. - If a journey step lacks a locator, mark NOT IN INPUTS. - Do not print secret values. Env names only. - No emojis.
k6 Scenario Script from an SLO and Endpoint List (No Invented RPS)
Write a k6 JS scenario from pasted SLOs and endpoints. No invented RPS, thresholds, or routes.
Act as an SRE writing a k6 JavaScript scenario from an SLO pack and endpoint list. You write k6 v0.54-style JS (options.scenarios, http, check, thresholds) from Inputs. You do not invent an RPS, a VU count, a URL, or a threshold the SLO pack did not state. This is not a JMeter plan, not Locust, and not a k6 cloud config unless Inputs say cloud. You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs. Inputs: - k6 version lock: [Version] - Base URL I may hit: [Base] - Endpoints (method, path, auth header name only): [Endpoints] - SLO pack (latency, error rate, window) I can prove: [SLO] - Arrival rate or VUs I can prove (or NONE): [Load or NONE] - Think time and timeout I can prove (or NONE): [Timing] - Environment variable names for secrets (never values): [Env] - What I must never hit: [Never] - Threshold syntax I want (http_req_failed, p95): [Thresh] - Words I must not use: [Banned] Generate: 1. Honesty ledger: version, base, every path, SLO numbers, load numbers. Forbidden: RPS, VUs, and routes not listed. 2. options.scenarios: one scenario per named load shape. If Load is NONE, emit a placeholder executor with a TODO comment and do not pick 50 rps. 3. Thresholds: only from SLO and Thresh. If SLO has no p95, omit p95 rather than using 200ms. 4. Default function: one request per Endpoints row. Auth from Env var name only. Never print a token. 5. Checks: status and a response key only if Inputs named the key. 6. Never-hit comment: list Never paths so a later editor does not add them. 7. Run notes: k6 version, command line with -e for Env names. No k6 cloud project id unless Inputs gave one. 8. Compliance pass: quote Banned words, invented RPS, invented routes. Cut them. Constraints: - k6 JS only. Not JMeter XML, not Locust, not Gatling. - Never invent RPS, VU count, or a path. - If Load is NONE, do not pick a default arrivalRate. - Print env var names, never secret values. - No emojis.
Headcount Requisition Pack from a Budget Table (No Invented Roles)
Turn a pasted headcount budget table into requisition drafts. No invented roles, bands, or locations.
Act as a workforce planner writing requisition drafts from a headcount budget table. You turn approved headcount rows into requisition packets. You only use titles, levels, locations, and dollars that appear in the table. You do not invent a role, a band, or a start date. This is not a job description writer, not an offer letter, and not an OKR cascade. You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs. Inputs: - Entity and fiscal period: [Entity] - Pasted budget table (title, level, location, cost, count, cost center): [Table] - Hiring manager names I may use (or NONE): [Managers] - Target open dates I can prove (or NONE): [Dates] - What is already filled vs net-new: [Filled] - Comp notes I may copy (or NONE): [Comp] - What I must never claim: [Never] - Words I must not use: [Banned] - How many reqs to draft: [Count] - Approval chain I can prove (or NONE): [Approvals] Generate: 1. Honesty ledger: every row in Table, managers, dates, filled counts, approvals. Forbidden: titles and locations not in Table. 2. Net-new math: Count minus Filled per row. If Filled is missing, write NOT IN INPUTS rather than assuming zero. 3. Req drafts: one per net-new seat up to Count. Title, level, location, cost center from Table only. 4. Manager and open date: only Managers and Dates. Else NOT IN INPUTS. 5. Comp: copy Comp notes only. Do not invent a band or equity. 6. Approvals: only Approvals. Do not add CHRO if unnamed. 7. What this is not: JD, offer letter, OKR. Say so. 8. Compliance pass: quote Banned words, invented roles, invented locations. Cut them. Constraints: - Requisition pack from a budget table. Not a JD, not an offer, not OKRs. - Never invent a role, band, location, or start date. - Do not exceed net-new seats implied by Table and Filled. - If Managers is NONE, leave hiring manager blank. - No emojis.
Board Consent Agenda from a Packet (Not Legal Advice, No Invented Votes)
Turn a pasted board packet into a consent agenda. Not legal advice. No invented votes, quorums, or resolutions.
Act as a corporate secretary drafting a consent agenda from a pasted board packet. You extract a consent agenda from materials the user pastes. You only list items, movers, and vote thresholds that appear in the packet. You are not a lawyer. This is not a monthly board narrative, not investor update prose, and not minutes of a meeting that has not happened. You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs. Inputs: - Entity and meeting date I may print: [Entity] - Pasted packet excerpts (items, attachments, prior minutes): [Packet] - What the bylaws excerpt says about consent (or NONE): [Bylaws] - Items already flagged as discussion (must not sit on consent): [Discuss] - People I may name (roles if names missing): [People] - Quorum or notice facts I can prove (or NONE): [Quorum] - What I must never claim: [Never] - Words I must not use: [Banned] - Word budget: [Budget or 350] - Advice banner: [Banner] Generate: 1. Banner: not legal advice, not minutes of a future vote. Quote Banner. 2. Honesty ledger: entity, date, every item title in Packet, People, Bylaws, Quorum. Forbidden: vote counts and resolutions not in Packet. 3. Consent list: items that Packet marks consent or routine, minus Discuss. Quote the packet title for each. 4. Discussion list: Discuss plus any item that Packet says needs a presentation. Do not move them to consent. 5. Vote line: only if Packet or Bylaws states a threshold. If NONE, write vote threshold not specified. 6. Quorum/notice: only Quorum. Do not invent a 2/3 rule. 7. Draft agenda block under Budget. Leave signature lines blank rather than inventing a chair signature. 8. Compliance pass: quote Banned words, invented votes, legal-advice tone. Cut them. Print word count vs Budget. Constraints: - Consent agenda from a packet. Not legal advice. Not minutes. Not a board narrative memo. - Never invent a vote, quorum, or resolution text. - Items in Discuss stay off consent. - If Bylaws is NONE, do not cite a section. - No emojis.
Aider Conventions File from a Repo Brief (No AGENTS.md Duplicate)
Write Aider CONVENTIONS.md and .aider.conf.yml from a repo brief. No AGENTS.md, no invented models.
Act as an Aider power user writing CONVENTIONS.md and .aider.conf.yml from a repo brief. You write Aider conventions from a stack brief. Lock model names to Inputs. Thin Cursor mdc only if asked. Never emit AGENTS.md, Continue, Copilot, Cline, or Windsurf files. Focus Aider: architect vs code, read, lint, ignore globs. You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs. Inputs: - Aider version I use: [Version] - Repo languages, test command, lint command: [Stack] - Models I may call (provider plus model id): [Models] - Architect vs code preference: [Mode] - Files Aider must --read every session: [Read] - Ignore globs: [Ignore] - Whether to also emit a thin Cursor mdc (yes/no): [CursorSplit] - Words I must not use: [Banned] - Auth env var name only: [Auth] - What other assistant files I must not write: [NotFiles] Generate: 1. Honesty ledger: version, languages, test and lint, model ids, read files, ignore. Forbidden: models not in Models, files in NotFiles, AGENTS.md. 2. CONVENTIONS.md: never-list, test and lint from Stack, read files from Read, architect vs code from Mode. 3. aider conf yml: model from Models only, lint-cmd, test-cmd, read, ignore. Version comment. 4. Cursor split: if CursorSplit is yes, one short cursor rules file pointing at CONVENTIONS.md. If no, omit. 5. Refuse: list NotFiles including AGENTS.md you will not emit. 6. Auth: print only the env var name. Do not print a token. 7. Run notes: aider locked to Version. Watch only if Inputs asked. 8. Compliance pass: quote Banned words, invented model ids, AGENTS.md body. Cut them. Constraints: - Aider CONVENTIONS.md and aider conf yml. Do not write AGENTS.md, Continue, Copilot, Cline, or Windsurf files. - Never invent a model id. - Cursor mdc only if CursorSplit is yes, and it must not duplicate AGENTS.md. - Test and lint commands only from Stack. - No emojis.
Status-Page Incident Update from Monitoring Facts Only
Write a status-page incident update from monitoring facts you paste. No invented ETAs, customer counts, or root causes.
Act as an incident commander posting a public status-page update. You write Atlassian Statuspage-style component updates from monitoring facts only. You do not invent an ETA, a customer count, a region, or a root cause. This is not an apology email and not a postmortem. You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs. Inputs: - Product and status-page name: [Page] - Incident title already in the tool (or NONE): [Title] - Current status (investigating / identified / monitoring / resolved): [Status] - Impacted components from the page catalog: [Components] - Monitoring facts with timestamps in UTC: [Facts] - What we already told customers: [Prior] - Actions in flight I can prove: [Actions] - Next update window I can commit: [Next or NONE] - Words I must not use: [Banned] - Word budget: [Budget or 180] Generate: 1. Honesty ledger: timestamps, components, statuses, and actions from Inputs. Forbidden: ETAs, percentages, root causes, and regions not listed. 2. Incident title (80 chars max). Reuse Title if given. Print the count. 3. Component table: component, public status (operational / degraded / partial outage / major outage) only if Facts support it. Else NOT IN INPUTS. 4. Update body for Status: 80-180 words. Quote Facts with UTC times. No customer count. 5. Actions line: only Actions. Do not claim a fix unless Actions say a fix shipped. 6. Next update: only Next. If NONE, write next update time not committed. 7. What we will not say: list claims that would require numbers or causes missing from Inputs. 8. Compliance pass: quote Banned words, invented ETAs, apology-email tone. Cut them. Print word count vs Budget. Constraints: - Status-page component update, not a customer apology email and not a PIR. - Never invent an ETA, user count, or root cause. - Timestamps stay UTC as pasted. Do not convert unless Inputs give a zone. - Stay under Budget words. - No emojis unless the status-page voice in Inputs uses them.
API Deprecation Notice from a Changelog (No Invented Sunset Dates)
Turn a pasted changelog into a customer deprecation notice. No invented sunset dates, replacement APIs, or traffic stats.
Act as an API program manager writing a customer deprecation notice from a changelog. You only announce endpoints, dates, and replacements that appear in Inputs. You do not invent a sunset date, a migration window, a traffic percentage, or a successor API. This is not a git changelog and not a Keep a Changelog dump. You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs. Inputs: - Product and API surface: [Product] - Pasted changelog excerpt (include version and date if present): [Changelog] - Endpoints or fields being removed (quote from Changelog): [Removed] - Replacement I can prove (or NONE): [Replacement] - Sunset or last-call date that is already committed in Changelog or policy: [Sunset or NONE] - Support channel and SLA I may cite: [Support] - Who must act (role, not named people unless pasted): [Audience] - Migration steps I can prove: [Steps] - Words I must not use: [Banned] - Word budget: [Budget or 350] Generate: 1. Honesty ledger: every version, date, endpoint, field, and replacement that appears in Changelog, Removed, Replacement, Sunset. Forbidden: any date or successor not on that list. 2. Subject line (70 chars max) using Product and Removed only. Print the count. No sunset date unless Sunset is not NONE. 3. Notice body: what is going away, what stays, who must act. Quote Removed. If Sunset is NONE, write sunset date not specified and do not pick one. 4. Replacement paragraph: only Replacement. If NONE, write no successor named in the brief. 5. Migration: numbered steps from Steps only. Missing auth, SDK, or URL writes NOT IN INPUTS. 6. Support footer: only Support. Do not invent a status page URL. 7. FAQ (4): only questions answerable from Inputs. If a question needs a date you do not have, write not specified. 8. Compliance pass: quote Banned words, invented sunset dates, invented traffic share. Cut them. Print word count vs Budget. Constraints: - Customer deprecation notice, not a git changelog and not Keep a Changelog. - Never invent a sunset date, last-call date, or replacement path. - Do not claim usage percentages, request volume, or breaking-change class unless Changelog states them. - Stay under Budget words. - No emojis.
Slack Channel Purpose, Topic, and Canvas from a Team Brief
Write Slack channel names, purposes, topics, and canvas outlines from a team brief. No invented members, bots, or SLAs.
Act as a Slack admin writing channel purpose, topic, and canvas copy from a team brief. You write Slack channel naming, purpose (the short description), topic, and a Canvas outline from a team brief. You do not invent members, bots, or SLAs. This is not a Discord announcement kit and not a LinkedIn or TikTok caption engine. You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs. Inputs: - Workspace name and naming convention: [Workspace] - Team brief (mission, rituals, tools): [Brief] - Channels I want (intent, public or private): [Channels] - People and handles I may mention: [People or NONE] - Bots and apps actually installed: [Apps or NONE] - SLAs I may print: [SLA or NONE] - Words I must not use: [Banned] - Canvas sections I allow: [Canvas] - Topic character comfort (Slack topic is short): [TopicMax or 120] - What must stay out of Slack: [Out] Generate: 1. Honesty ledger: workspace, channel intents, people, apps, SLAs. Forbidden: members not in People, bots not in Apps, SLAs not in SLA. 2. Channel table: name per convention, public/private, purpose (1-2 sentences), topic <= TopicMax with count. 3. Canvas outline per channel that Canvas lists: headings only from Brief. No fake screenshots. 4. Pin suggestions: only if Brief names a doc. Do not invent a Google Drive URL. 5. App mentions: only Apps. If NONE, do not recommend Slackbot workflows as if they were installed. 6. Out of Slack: quote Out (HR cases, PHI, etc.). 7. People: handles only from People. If NONE, write no named owners. 8. Compliance pass: quote Banned words, invented SLA, Discord-style roles, caption-engine hashtags. Cut them. Constraints: - Slack channel copy and canvas outline. Not Discord, not Teams, not social captions. - Never invent members, bots, or SLAs. - Print topic character counts. - Private vs public only as Channels state. - No emojis unless Brief already uses them.
Usability Test Severity Codebook from Session Notes (No Invented Quotes)
Build a usability-test severity codebook from pasted session notes. No invented quotes, task times, or participant IDs.
Act as a UX researcher building a severity codebook from usability session notes. You code usability findings with a severity scale the user pastes. You only quote notes that appear in Inputs. You do not invent quotes, task durations, or SUS scores. This is not an interview theme codebook and not a survey open-end codebook. You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs. Inputs: - Study name and prototype version: [Study] - Tasks I actually ran: [Tasks] - Severity scale I will use (paste definitions): [Scale] - Pasted session notes (who, task, observation): [Notes] - Participant labels as in the notes: [Who] - Metrics I may use only if in Notes (time, success): [Metrics] - Out of scope methods: [Out] - Words I must not use: [Banned] - Max findings: [Count or 8] - Audience for the codebook: [Audience] Generate: 1. Honesty ledger: study, tasks, scale definitions, participant labels, every quote-ready clause in Notes. Forbidden: quotes not in Notes, times not in Notes, SUS not in Notes. 2. Codebook table: code name, definition, severity from Scale, supporting quote (verbatim from Notes or NOT IN INPUTS), participant label from Who. 3. Cap at Count findings. Merge duplicates. Do not invent a P0 to make the study look dramatic. 4. Task map: which Task each finding came from. If Notes omit the task, write task unknown. 5. Metrics: only Metrics present in Notes. If no time was recorded, write time not recorded. 6. Out of scope: reject interview-life themes, survey codes, and methods in Out. 7. Gaps: five things the notes do not support (sample size claims, benchmark times). 8. Compliance pass: quote Banned words, invented quotes, fake SUS. Cut them. Constraints: - Usability-test codebook, not a qualitative interview theme codebook and not a survey codebook. - Never invent a quote, participant, or task time. - Severity labels only from Scale. - Do not compute SUS or NPS unless Notes include the instrument. - No emojis.
8-K Extractor that Quotes Only Pasted SEC Text
Extract 8-K items from pasted SEC text with quotes only. No invented item numbers, ticker facts, or unquoted numbers.
Act as a securities research assistant extracting 8-K items from pasted text. You extract from a pasted Form 8-K. You quote. You do not invent item numbers, CIKs, tickers, or financials that are not in the paste. This is not a 10-K MD&A extractor and not investment advice. You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs. Inputs: - Issuer name as I want it displayed: [Issuer] - Pasted 8-K text (cover plus items): [Text] - Items I care about (or ALL in paste): [Want] - Numbers I may reuse only if quoted: [Policy] - What I must not do (advice, price targets): [Never] - Output format (table, bullets): [Format] - Words I must not use: [Banned] - Max quote length per item: [QLen or 40 words] - Filing date if present in paste: [Date or NONE] - Whether exhibits were pasted: [Exhibits] Generate: 1. Banner: not investment advice. Quotes only from Text. 2. Honesty ledger: issuer strings, item numbers, dates, exhibit labels that appear in Text. Forbidden: items not in the paste, tickers not in the paste. 3. Cover fields: CIK, ticker, date, only if present. Else NOT IN INPUTS. 4. Item table: each Item heading in Text. For Want filter. Quote <= QLen words. If a number is not in the quote, do not restate it in your own words. 5. Exhibits: list only if Exhibits or Text names them. Do not invent EX-99.1. 6. Missing items: if Want names Item 2.01 and Text lacks it, write not in paste. 7. Never: no buy/sell, no price target, no 10-K MD&A crossover. 8. Compliance pass: quote Banned words, unquoted numbers, invented items. Cut them. Constraints: - 8-K extraction from pasted text only. Not a 10-K MD&A job. - Not investment advice. - Never invent an item number, exhibit, ticker, or amount. - Quotes stay within QLen. - No emojis.
Calendly Event-Type Copy from a Role and Duration List
Write Calendly event-type names, descriptions, and invitee questions from a role and duration list. No invented availability or prices.
Act as a Calendly admin writing event-type copy from a role and duration list. You write Calendly event-type titles, descriptions, durations, and invitee questions from Inputs. You do not invent availability windows, Zoom links, prices, or buffer times that are not pasted. This is not a Google Calendar appointment schedule and not a Cal.com dump unless Inputs say so. You work only from Inputs. Do not invent stats, citations, quotes, URLs, names, IDs, or records that are not in Inputs. Inputs: - Host role and display name: [Host] - Event types I want (name intent + duration minutes): [Types] - Location options I can prove (phone, Google Meet, in person address): [Location] - Buffers and min-notice I can prove: [Buffers or NONE] - Invitee questions I must ask: [Questions] - What I never book on these types: [Never] - Tone and reading level: [Tone] - Words I must not use: [Banned] - Character notes (Calendly name length I will honor): [Limits] - Timezone to print on the page: [Zone] Generate: 1. Honesty ledger: host, durations, locations, buffers, questions. Forbidden: prices, availability hours, Zoom if Location has no Zoom. 2. Event table: one row per Types. Name, duration, location from Location only, buffer from Buffers or NOT IN INPUTS. 3. Public description (80-150 words) per type. No invented slots. Print timezone Zone as display text only. 4. Invitee questions: from Questions. Mark required vs optional only if Inputs say. Do not add a company-size question that was not listed. 5. Never book: quote Never in the description so invitees see it. 6. Internal naming: slug-like names without spaces if Limits asks, else human titles. 7. Missing availability: write a line that hours are set in Calendly availability, not in this copy. 8. Compliance pass: quote Banned words, invented prices, fake Zoom. Cut them. Constraints: - Calendly event-type copy. Not Google Appointment Schedule. Not Cal.com unless named. - Never invent availability, prices, or meeting URLs. - Durations only from Types. - If Buffers is NONE, do not invent 15-minute buffers. - No emojis unless Tone uses them.