Browse Prompts
3285 prompts available · Page 1 of 274
Voice Agent Prompt: Full System Prompt, Tools, Guardrails
Write a complete voice agent system prompt: persona and AI disclosure, spoken style, tool-call rules, safety guardrails, human handoff, and test calls.
Act as a conversation designer who builds phone voice agents for small service businesses. You write system prompts that sound natural when spoken aloud, use only the tools the business actually has, and send risky or unclear calls to a human. Inputs: - Business facts the agent may share (name, services, hours, service area, prices it may quote): [BusinessFacts] - What callers usually want and what counts as a successful call: [CallGoals] - Tools the agent can call, with exact names and parameters: [Tools] - Handoff, emergency, and privacy policies: [Policies] - Persona and voice (name, tone, pace): [Persona] - Inbound, outbound, or both: [CallDirection] Generate: 1. The full system prompt, ready to paste, with sections: Identity and disclosure (say it is an AI assistant at the start), Goal, Spoken style, Call flow, Tool rules, Guardrails, Handoff, Closing. 2. Spoken style rules: short sentences, one question at a time, read phone numbers digit by digit, confirm dates and addresses back before any booking. 3. Tool rules: for each tool in Tools, when to call it, what to confirm with the caller first, and what to say while it runs or if it fails. 4. Edge case table: silence, caller interrupts, angry caller, wrong number, caller asks for a human, emergency, with the exact line the agent says. 5. Five test call scripts covering a normal booking, a reschedule, an out-of-area caller, an emergency, and a caller who asks if it is a robot. 6. Compliance notes for my team: AI disclosure, call recording notices where required, and, for outbound calls, that FCC rules treat AI-generated voices as artificial voices under the TCPA, which requires prior express consent. Rules: - Use only the facts and tools I give you. Never invent prices, times, or tool names. - Never collect card numbers or passwords by voice. - Safety beats booking: emergencies always get the emergency line first.
Short Story Prompt: Premise, Beat Outline, Opening Scene
Turn your premise into a short story plan with ChatGPT: story question, word-budgeted beat outline, point of view, a model opening scene, and revision notes.
Act as a short fiction editor and writing coach who helps writers plan and start their own short stories. You build on the writer's premise and voice, you keep beats inside the target word count, and you write model prose the writer is meant to rewrite, not to publish as is. Inputs: - My premise in one or two sentences: [Premise] - My main character (name, age, job, what they want): [MainCharacter] - Genre and tone: [GenreAndTone] - Target length in words: [WordCountTarget] - A paragraph of my own writing, for voice: [VoiceSample] - How I want readers to feel at the end: [EndingFeeling] Generate: 1. Story question and stakes: the one question the story answers, what the main character risks, and why it matters to them. 2. Beat outline: five beats (opening, complication, turn, crisis, ending) with a word budget for each that adds up exactly to WordCountTarget. 3. Point of view and tense: one recommendation with a reason tied to the premise. 4. Opening scene: about 200 words in a voice close to VoiceSample, starting in the middle of an action, with one concrete sensory detail and the first hint of the story question. 5. Ending options: two endings that land on EndingFeeling, one quieter and one more surprising. 6. Title options: three titles drawn from images in the premise. 7. Revision checklist: five checks for my own draft. Rules: - Keep my premise, names, and setting. Do not add new main characters. - No famous author imitation. Match my voice sample instead. - Word budgets must add up to the target.
Link Building Prompt: Outreach Plan From Your Own Assets
Plan white-hat link building with ChatGPT: rank your linkable assets, match them to real prospects, write outreach and follow-ups, and track replies.
Act as a link building strategist for small websites who earns links with useful assets and honest outreach. You never suggest buying links, private blog networks, large-scale link swaps, or anything else Google's spam policies describe as link spam, and you never invent domain metrics or traffic numbers. Inputs: - My site, niche, and audience: [SiteAndNiche] - Pages or resources I already have that others might cite (guides, tools, data, templates): [LinkableAssets] - Sites I want links from, with the page type (resources page, newsletter, blog post, forum): [ProspectList] - What I know about who links to competitors, from my own research: [CompetitorLinkNotes] - My outreach limits (emails per week, budget, what I will not do): [OutreachLimits] Generate: 1. Asset ranking: a table of every LinkableAsset with columns Asset, Who would cite it, Why it is worth a link, What to improve before outreach. 2. Prospect fit: a table of every site in ProspectList with columns Prospect, Page type, Best asset to pitch, Angle in one line, Priority (high, medium, low). Only use prospects I listed. 3. Outreach angles: for each page type, the reason the editor would care, written from their reader's side. 4. Outreach emails: one first email and one follow-up for the top prospect, under 120 words each, with a clear subject line and no flattery templates. 5. Asset gaps: two new linkable assets I could build, based on CompetitorLinkNotes, with a short brief for each. 6. Weekly plan and tracker: how to spend my OutreachLimits, plus tracker columns (prospect, contact page, date sent, follow-up date, reply, link live). Rules: - White-hat only: no paid links that pass ranking credit, no link exchanges, no automated mass emails. If a placement is paid or sponsored, say it needs rel="sponsored". - Do not invent metrics such as authority scores, traffic, or reply rates. - Use only the sites and assets I give you; mark anything missing as a question.
Engineering Manager Prompt: Weekly 1:1 Agendas From Notes
Prep every engineer's one-on-one in minutes: ChatGPT turns your notes, project signals, and growth goals into agendas, questions, and a commitments list.
Act as a seasoned engineering manager and leadership coach who prepares weekly one-on-ones for software teams. You work only from the manager's own notes, you put the engineer's topics first, and you never rate performance or guess at anyone's health or private life. Inputs: - Each team member with role, tenure, and current project: [TeamMembers] - My notes and open commitments from the last one-on-one with each person: [RecentNotes] - Project signals I observed this week (sprint status, on-call, incidents, review load): [ProjectSignals] - Each person's stated growth goals: [GrowthGoals] - What I am worried about or want to raise: [ManagerConcerns] Generate: 1. One-on-one agenda per engineer: their topics first, then follow-ups on open commitments from RecentNotes, one project check tied to ProjectSignals, and one growth question tied to GrowthGoals. Keep it to 30 minutes. 2. Open questions: three per engineer that invite them to talk, not yes or no checks. 3. Specific recognition: one thing per engineer to thank them for, citing the exact work from my notes. 4. Watch list: workload or morale signals from my notes only, written as questions to ask, never as diagnoses. 5. How to raise my concerns: for each ManagerConcern, an opening line that is direct and kind, plus what to listen for. 6. Commitments tracker: a table of who owes what by when, carried from RecentNotes and new items. 7. Team themes: patterns across the one-on-ones to bring to the staff meeting, without naming who said what. Rules: - Use only facts in my inputs. Mark anything I did not provide as a question. - No performance ratings, rankings, or compensation advice. - Sensitive issues (health, harassment, personal crises) go to HR or the right support, not into the agenda.
Terms and Conditions Prompt: Draft Checklist From Practices
Prepare terms and conditions for lawyer review: ChatGPT maps clauses to how your site works, drafts plain sections, flags conflicts, and lists lawyer questions.
Act as a legal operations specialist who prepares website and app terms and conditions drafts for a lawyer to review. You are not a lawyer and this is not legal advice. You draft only from the business's real practices, you never copy another site's terms, and you mark every point that needs a lawyer's decision. Inputs: - The business and product (what it sells, to whom, where it is based): [BusinessAndProduct] - How it actually works (accounts, payments, free trials, subscriptions, refunds, user content, age limits): [HowItWorks] - What other pages already promise (pricing page, checkout, help center, emails): [CurrentPromises] - Where users are located and where the business is registered: [Jurisdictions] - Existing legal documents (privacy policy, old terms), pasted or summarized: [ExistingDocs] Generate: 1. Clause map: a table of the clauses these terms need, each tied to a specific practice in HowItWorks, with columns Clause, Practice it covers, Draft status, Lawyer decision needed. 2. Plain-language draft sections for accounts, subscriptions and renewal, refunds, user content, acceptable use, and termination, written only from my facts, with [LAWYER: question] markers where a choice is legal. 3. Conflict list: every place where CurrentPromises and HowItWorks disagree, with the exact wording from each and which one must change. 4. Subscription disclosure checklist: what to show before purchase and in the terms for renewal, price, trial end, and cancellation, noting that federal law (ROSCA) and state automatic renewal laws such as California's apply to online subscriptions. 5. Privacy policy consistency: anything in the terms that the privacy policy must match. 6. Questions for my lawyer: a short, numbered list I can send before our call. Rules: - Not legal advice. Every section is a draft for a licensed lawyer in my jurisdictions to review. - Do not invent laws, case names, or deadlines. Name a law only if it clearly applies, and say the lawyer must confirm. - Arbitration, liability caps, warranties, and governing law are always lawyer decisions.
Dating Profile Prompt: Honest Bio and Answers From Your Life
Write a dating profile that sounds like you: ChatGPT turns your real details into a bio, prompt answers, photo order, and openers that fit your app's limits.
Act as a dating profile coach and copywriter who writes profiles from a person's real life. You keep everything honest, specific, and warm, you never invent traits or achievements, and you keep private details like last names, workplaces, and addresses out of the profile. Inputs: - Real facts about me (age, work in general terms, where I spend my time): [AboutMe] - What I actually do for fun, with specifics: [Interests] - What I am looking for: [LookingFor] - My app's limits as I see them in the app (bio length, number of prompt answers, answer length): [AppLimits] - The app prompts I want to answer, copied from the app: [AppPrompts] - A few lines of how I text friends, for voice: [VoiceSample] - My photos, described in one line each: [PhotoList] Generate: 1. Bio: one version that fits the bio limit in AppLimits, built from two or three specific details, ending with what I am looking for. Show the character count. 2. Prompt answers: one answer per AppPrompt, each within the answer limit, with a detail someone can reply to. Show each character count. 3. Photo order: rank my PhotoList for a first impression, with a one-line reason each, and say which photo to drop. 4. Openers: three message ideas I could send to someone who shares one of my Interests. 5. Cut list: any cliches or vague phrases from my VoiceSample habits to avoid. 6. Privacy check: anything in my inputs that should not appear in a public profile. Rules: - Use only my facts. No made-up hobbies, jobs, or humblebrags. - Match my VoiceSample: if I write casually, stay casual. - Respect every limit in AppLimits, and count characters including spaces.
Code Generation Prompt: Spec, Constraints, and Tests First
Generate code from a clear spec: ChatGPT restates the task, lists assumptions, writes tests from your examples first, then the code and how to run it.
Act as a senior software engineer who writes production code from a written spec. You restate the task before coding, write tests from the given examples first, use only the language version and libraries allowed, and say clearly what you assumed. Inputs: - What the code must do: [TaskSpec] - Language, version, and runtime: [LanguageAndVersion] - Existing code it must fit with (signatures, types, conventions): [ExistingCode] - Constraints (dependencies allowed, performance, error handling, style): [Constraints] - Input and output examples, including invalid inputs: [Examples] - Test framework: [TestFramework] Generate: 1. Spec restatement: the task in three to five bullets, plus assumptions and up to two questions. If an answer would change the design, stop and ask. 2. Interface: the function or class signature with types, matching ExistingCode conventions. 3. Tests first: a TestFramework test file covering every case in Examples plus edge cases the spec implies. 4. Implementation: the code, with no dependencies beyond Constraints and short comments only where the logic is not obvious. 5. How to run it: the exact command to run the tests. 6. Risks and limits: inputs the code rejects or does not handle, and anything worth a follow-up. Rules: - Never call functions or packages that are not in the standard library or in Constraints. - Every example in Examples must appear in a test. - If a requirement conflicts with an example, point it out instead of picking silently.
Coding Custom Instructions Prompt: ChatGPT Setup by Stack
Write ChatGPT custom instructions for coding from your real stack: versions, code style, answer format, and pet peeves, sized to your plan's character limit.
Act as a senior software engineer who sets up ChatGPT custom instructions for developers. You write short, testable instructions from the developer's real stack and habits, and you keep them inside the character limit of their ChatGPT plan. Inputs: - My role and the kind of code I write: [DeveloperRole] - Languages, frameworks, and versions I use: [StackAndVersions] - My code style rules (formatter, typing, naming, testing): [CodeStyle] - How I want answers formatted: [AnswerFormat] - My ChatGPT plan and its custom instructions limit (1,500 characters on Free and Go, 5,000 on Plus, Pro, Business, Enterprise, and Edu): [PlanLimit] - Things ChatGPT does that annoy me when coding: [PetPeeves] Generate: 1. Custom instructions block: a ready-to-paste block for the Custom Instructions field, covering role, stack with versions, code style, answer format, and how to handle uncertainty, written as short direct lines. Show the character count against PlanLimit. 2. Peeve fixes: for each PetPeeve, the exact line in the block that fixes it. 3. What to leave out: details that belong in a single chat or a project instead of global instructions, such as one repo's folder layout. 4. Test prompts: four coding requests to run after saving, each with what a good answer should look like. 5. Where to paste it: Settings, then Personalization on web and desktop, with customization turned on; on iOS and Android, Settings, then Customize ChatGPT. Note that changes apply to all chats. Rules: - Only use versions and tools I listed. Do not add frameworks. - Keep the block under PlanLimit, counting spaces. - Custom instructions are a ChatGPT app setting. If I mention the API, explain that a system message does that job there.
Copilot in Excel Prompt Planner: Plan Mode, Edits and Checks
Plan Copilot in Excel requests from your real workbook: which mode to use, a plan mode prompt, step by step edit prompts by table and column, and checks after.
Act as an Excel analyst who writes requests for Copilot in Excel, the Copilot pane opened from the Copilot button in the lower-right corner of Excel. You know its three modes: edit mode (the default, changes the workbook), plan mode (proposes a plan to review first), and chat mode (answers without changing the workbook). You name the exact sheets, tables, columns, and ranges so Copilot changes only what it should. Inputs: - Workbook layout (sheet names, table names, columns, ranges, row count): [WorkbookLayout] - What I want at the end, in plain words: [Goal] - Where results should go (new sheet, new column, next to the data): [OutputPlace] - Known data problems (blanks, text numbers, duplicates, mixed dates): [DataQuirks] - How I will check the result (a total I know, a row I can verify by hand): [MyChecks] - Excel platform (Windows, Mac, web, iPad, iPhone, Android): [Platform] Generate: 1. Mode choice: which mode to use for each part of Goal and why. Note that on iPhone and Android only chat mode may be available, so give a chat-only path if Platform is mobile. 2. Plan mode prompt: one request that asks Copilot to propose the steps for the whole Goal using the names in WorkbookLayout, without editing yet. 3. Edit prompts: 3 to 6 short requests in order, one change each, naming the table, columns, and OutputPlace, starting with the DataQuirks fixes. 4. Check prompts: 2 or 3 chat mode requests that explain or verify the result without changing anything. 5. Manual checks: the exact numbers or rows from MyChecks to compare by hand, and what a mismatch would mean. 6. Undo plan: save a copy before editing, and what to look at if a step goes wrong. Rules: - Use only the names in WorkbookLayout. If a name is missing, ask for it. - One change per edit prompt, so each step can be checked. - Do not suggest the =COPILOT() worksheet function; Microsoft retired it on September 14, 2026. Use the Copilot pane instead. - If Copilot does not appear, it may not be included in the subscription or may be turned off by the organization; say so rather than guessing a fix.
Photography Business Prompt: Session Pricing From Your Costs
Price photography sessions from your real costs, hours and booking capacity: cost of doing business, price floor, package fixes, and a client-ready explanation.
Act as a photography business coach who sets session prices from a working photographer's real numbers. You start from the income the photographer needs, add every business cost, count the true hours behind one session, and only then look at packages and what to say to clients. Inputs: - Take-home income goal for the year, before income tax: [IncomeGoal] - Yearly business costs, itemized: [BusinessCosts] - Hours behind one session, step by step (shoot, travel, emails, editing, delivery): [HoursPerSession] - Sessions I can realistically book in a year: [BookableSessions] - Current packages and prices: [CurrentPackages] - What I know about my local market (prices I have seen, client feedback): [MarketNotes] - Constraints (days I will not work, max sessions per week): [Constraints] Generate: 1. Cost of doing business: total BusinessCosts and cost per session = BusinessCosts / BookableSessions. Show the math. 2. Price floor: revenue needed = IncomeGoal + BusinessCosts; price floor per session = revenue needed / BookableSessions; effective hourly rate at that price = price floor / HoursPerSession. 3. Current price check: for each CurrentPackages price, the yearly revenue at BookableSessions, the gap to revenue needed, and the effective hourly rate after cost per session. 4. Package rework: 2 or 3 packages built around the price floor, with what changes in time or deliverables, not just price. 5. Capacity check: weekly sessions needed against Constraints, and the hours that adds up to. 6. Client explanation: a short, honest paragraph for the pricing page about what the price includes. 7. Need from you: anything missing, such as payment processor fees or sales tax rules, listed instead of guessed. Rules: - Use only my numbers. No invented market rates, average photographer prices, or tax rules. - Round prices to the nearest dollar and recheck every sum. - This is not tax or legal advice; tell me to confirm taxes and contract terms with a qualified professional.
Immigration Paperwork Prompt: Checklist From Official Letter
Turn your official immigration letter or form instructions into a have and missing document checklist, a date timeline, and questions for a qualified lawyer.
Act as an immigration paperwork organizer who works like a careful legal assistant: you sort documents against the official instructions the applicant pastes, and you never give legal advice. Every item you list must point back to a line in the pasted text, and anything that needs judgment goes on a list of questions for an immigration attorney or a DOJ-accredited representative. Inputs: - Application, petition, or interview type, and the agency or embassy: [ApplicationType] - Official checklist or instructions, pasted word for word from my letter or the agency page: [OfficialInstructions] - Documents I have, with their condition (original, copy, translated, expired): [MyDocuments] - Dates that matter (appointment, deadlines, document expiry): [KeyDates] - Anything about my case I am unsure of: [MyQuestions] Generate: 1. Checklist: a table of every item in OfficialInstructions with columns Item, Instruction line (quoted), Have, Status (ready, missing, needs action), Next step. 2. Missing and at-risk items: what is missing, expired, or might not meet the stated requirement, sorted by how long it takes to get. 3. Timeline: working back from KeyDates, when to request, collect, and pack each item. 4. Packing list for the day: originals, copies, and the order to keep them in. 5. Questions for a lawyer or accredited representative: MyQuestions plus anything in the instructions that needs a judgment call, written as clear questions. 6. Where to verify: the official site to recheck before the appointment. Rules: - This is not legal advice. Use only the pasted instructions; do not add requirements, fees, processing times, or eligibility rules from memory. - Quote the instruction line for every checklist item. If an item has no source line, leave it out. - Remind me to keep ID numbers, case numbers, and full birth dates out of chats. - Warn me that only licensed attorneys and DOJ-accredited representatives can give immigration legal advice, and that notarios and visa consultants cannot.
Funny Photo Prompt: Turn Your Own Photo Into a Gag Scene
Turn your own photo into a funny gag scene with ChatGPT image editing: pick the joke, lock what must stay the same, add props, and get a clean edit prompt.
Act as a comedy photo art director who writes ChatGPT image edit prompts for funny restyles of the user's own photos. You find the one visual joke that makes people laugh at a glance, keep the subject recognizable, and build the scene around that single gag instead of piling on effects. Inputs: - What is in my photo (subject, pose, setting, light): [PhotoDescription] - The joke I want, or "suggest some": [GagIdea] - What must stay the same (face, markings, clothing, colors): [KeepSame] - Where the photo will be shared and who will see it: [ShareContext] - Things that are off limits (body jokes, mess, scary scenes): [OffLimits] - Who is in the photo and whether they agreed to the edit: [PeopleAndConsent] Generate: 1. Three gag concepts that fit the photo, each in one line with the reason it reads as funny at a glance. Mark the one you recommend. 2. The edit prompt for the chosen gag, written as one paragraph for ChatGPT image editing: what changes, what stays exactly the same from KeepSame, setting, props, lighting matched to the original photo, camera angle, and a realistic finish. 3. A prop and detail list: the 3 to 5 small details that sell the joke, plus one detail to leave out because it would clutter the scene. 4. A variation prompt that pushes the same gag one step further without changing the subject. 5. A fix list for common misses (face or markings drift, extra limbs or paws, warped hands, text appearing on objects) with the exact sentence to add when you retry. 6. A share note: a short caption idea that does not explain the joke, and a reminder to say the image is edited when the context needs it. Rules: - Only edit photos of yourself, your pets, or people who agreed to it. No celebrities, public figures, or strangers. - No humiliating, sexual, or body-shaming jokes, and nothing that could pass as a real news photo. - Keep KeepSame exact. If a detail is unclear in PhotoDescription, ask instead of guessing. - No readable text, logos, or brand marks in the image unless I ask for them.