These 53 prompts cover the week of a small or growing YouTube channel: deciding what the channel is for, finding video ideas in the questions your viewers ask, researching and fact-checking a script, writing hooks and endings, packaging a video, turning one video into Shorts, reading your YouTube Studio numbers, and handling sponsors and comments.
They are organised into seven sections, from channel positioning to sponsorships and monetisation. Several work from material you paste in, such as comments, transcripts and Studio figures, because the AI cannot see your channel. The analytics prompts compare your videos with your own history and say when a sample is too small to conclude anything. None of the prompts promise views, subscribers or income.
The packaging prompts aim to keep titles and thumbnails true to the video, and one prompt audits a title and thumbnail against what the video really delivers. Where a prompt involves a sponsor, a free product or an affiliate link, it puts the disclosure up front and points you to YouTube's current help pages and your local advertising rules, which change and which you should read yourself. Each prompt works in ChatGPT, Claude and Gemini. Fill in the {{VARIABLES}} with your own details and treat the output as a draft.
Channel Positioning & Planning
Most small channels stall because the channel has no clear promise, not because a single video failed. These prompts help you write a positioning statement from what you can credibly offer, choose a few content pillars, and find a gap without copying anyone. They finish with a 90-day plan sized to the hours you really have and a channel page that makes sense to a first-time visitor.
Best for: Getting a clear, testable channel promise before you plan more videos
You are a YouTube strategist who works with small channels. Help me write a positioning statement for my channel.
My channel topic: {{CHANNEL_TOPIC}}
Who I make videos for: {{TARGET_VIEWER}}
What I can credibly offer that others do not (experience, access, style): {{MY_EDGE}}
Channels I am often compared with: {{COMPARISON_CHANNELS}}
Rules:
- Base everything on what I wrote above. Do not invent credentials, subscriber numbers or results.
- Do not promise that any positioning will grow the channel. Treat it as a hypothesis to test.
- Avoid vague words like "valuable", "engaging" and "high-quality".
Output:
1. Three one-sentence positioning statements in the form "For [viewer] who [need], this channel [what it does] unlike [alternative] because [reason]".
2. For each, one sentence on the risk (who it excludes, what it makes harder).
3. The channel promise in 12 words or fewer, for a banner or bio.
4. A recommendation for which to test first, with the one thing I would watch over my next 10 videos to see whether it is working.
Fill in: CHANNEL_TOPIC TARGET_VIEWER MY_EDGE COMPARISON_CHANNELS
Fill in the four fields honestly, even if the answers feel small. Treat the three options as experiments.
Best for: Turning a loose topic area into a few sustainable pillars
You are a content planner for a small YouTube channel. Build content pillars from what I actually know and care about.
Channel positioning: {{CHANNEL_POSITIONING}}
Topics I can speak about without research: {{MY_TOPICS}}
Videos I can realistically make per month: {{VIDEOS_PER_MONTH}}
Formats I am willing to make (talking head, screen recording, voice-over, vlog): {{FORMATS}}
Rules:
- Propose 3 or 4 pillars, never more. Each must be narrow enough that a stranger could say what it is about.
- Do not assume a viral formula. Explain each pillar by the viewer need it serves.
- Keep the plan inside my stated capacity and formats.
- If two of my topics belong to different audiences, say so plainly rather than forcing them together.
Output a table with columns: Pillar name, Viewer need it serves, Example video format, Three example video ideas, Effort level (low/medium/high). Then give a suggested mix across {{VIDEOS_PER_MONTH}} videos a month and one sentence on what to drop if time runs short.
Fill in: CHANNEL_POSITIONING MY_TOPICS VIDEOS_PER_MONTH FORMATS
Be honest about capacity. A realistic plan you can keep beats an ambitious one you abandon.
Best for: Finding an angle in a crowded niche without copying anyone
You are a YouTube market analyst. I will give you notes about channels in my niche. Map the gaps I could serve.
My channel and niche: {{MY_CHANNEL_AND_NICHE}}
Notes on 3 to 6 channels (what they cover, formats, video titles you have seen, what viewers complain about in their comments): {{COMPETITOR_NOTES}}
My strengths and limits (budget, equipment, expertise, time): {{MY_STRENGTHS_AND_LIMITS}}
Rules:
- Work only from my notes. You cannot see live channel data, so do not state subscriber counts, view numbers or upload schedules. If I have not provided something, say it is unknown.
- Do not suggest copying another creator's videos, titles or thumbnails. Look for gaps in angle, depth, format, audience level or honesty.
- Mark each gap as "supported by my notes" or "a guess to verify".
Output:
1. A table of the channels with their apparent angle, level (beginner, intermediate, advanced) and format.
2. Five gaps, each with evidence from my notes, the audience it serves and whether my strengths fit.
3. The gap I should test first and a three-video test plan.
4. What I should check by watching the channels myself before deciding.
Fill in: MY_CHANNEL_AND_NICHE COMPETITOR_NOTES MY_STRENGTHS_AND_LIMITS
Watch the channels and take your own notes first. The AI can organise them but cannot look the channels up.
Best for: Planning repeatable formats that speed up your production
You are a series producer for educational and entertainment YouTube channels. Develop repeatable series concepts for my channel.
Channel topic and pillars: {{CHANNEL_TOPIC_AND_PILLARS}}
Viewer I am serving: {{TARGET_VIEWER}}
Production limits (time per video, equipment, solo or team): {{PRODUCTION_LIMITS}}
Series I have already tried and how they went: {{PAST_SERIES}}
Rules:
- Suggest 4 series ideas. Each must be repeatable at least 8 times without running out of material, and producible within my limits.
- Each series needs a consistent structure that I can reuse, such as a fixed question, constraint or format, so editing and scripting get faster over time.
- Be honest where a series idea depends on luck, access or a large audience.
- Do not predict performance.
For each series, give: a working name, a one-line premise, the repeatable structure in 4 steps, five episode ideas, what a viewer gets from watching three episodes in a row, and the main way it could fail. Finish by telling me which to pilot with a three-episode test and what I should compare afterwards.
Fill in: CHANNEL_TOPIC_AND_PILLARS TARGET_VIEWER PRODUCTION_LIMITS PAST_SERIES
Include honest notes about past series that fizzled. They tell the AI what to avoid.
Best for: Replacing an optimistic schedule with one you can sustain
You are a production manager for a solo or small-team YouTube channel. Create a 90-day plan I can actually keep.
Hours per week I can give to the channel: {{HOURS_PER_WEEK}}
My channel pillars: {{CONTENT_PILLARS}}
Current upload rhythm and what has been hard about it: {{CURRENT_RHYTHM}}
Goals for the period, stated as things I control (videos finished, tests run, not views): {{GOALS}}
Fixed dates I must work around: {{FIXED_DATES}}
Rules:
- Choose a rhythm I can sustain, and say what happens in a week I miss. Build in one buffer week per month.
- Break work into stages (idea, research, script, record, edit, package, publish, review) and estimate hours using my capacity, not an ideal.
- Do not promise results. Goals should be about process.
- If my hours do not support my current rhythm, say so and propose a smaller one.
Output: a month-by-month plan with videos, a weekly template showing which day does which stage, three experiments to run (one per month, each tied to a single change), and a monthly checkpoint question.
Fill in: HOURS_PER_WEEK CONTENT_PILLARS CURRENT_RHYTHM GOALS FIXED_DATES
Be realistic about hours. Rerun the prompt after a month with what actually happened.
Best for: Making your channel page clear to a first-time visitor
You are a copywriter for creators. Write the first things a new visitor sees on my channel.
Channel positioning and promise: {{CHANNEL_POSITIONING}}
What viewers will find (pillars, series, upload rhythm): {{WHAT_VIEWERS_FIND}}
Links and contact details I want to include (business email, website, socials): {{LINKS_AND_CONTACT}}
Tone I want (for example plain, dry, warm, curious): {{TONE}}
Rules:
- Use only facts I provide. Do not claim awards, numbers or a posting schedule I did not state.
- No filler such as "Welcome to my channel where I share amazing content". Be specific about the topic and the viewer.
- Keep the About text honest if I post irregularly.
- If I use affiliate links or take sponsorships, add a brief plain line saying so.
Output:
1. About section, under 900 characters, first two lines written to work before the "show more" cut.
2. Three banner or tagline options, each under 10 words.
3. A 45 to 60 second welcome or channel trailer outline: hook, who it is for, the three best things to watch, and a single ask.
4. A short list of anything missing from my inputs.
Fill in: CHANNEL_POSITIONING WHAT_VIEWERS_FIND LINKS_AND_CONTACT TONE
Put your strongest sentence in the first two lines of the About text, since many viewers never expand it.
Video Ideas From Real Audience Signals
Ideas work best when they come from questions people ask, not from guessing at trends. Here the AI organises the comments, questions and search suggestions you collect yourself and builds a picture of who is watching, scores your idea list against your own priorities, and pulls stories from your experience. One prompt covers timely news videos without spreading mistakes, and one plans collaborations that give the other creator a reason to agree.
Best for: Building a backlog from what viewers genuinely ask
You are a video idea strategist for a small YouTube channel. Turn real questions into video concepts.
Channel topic: {{CHANNEL_TOPIC}}
Questions people have asked (from comments, emails, forums, DMs, or questions you copied from search suggestions): {{AUDIENCE_QUESTIONS}}
Viewer level (beginner, intermediate, advanced): {{VIEWER_LEVEL}}
Rules:
- Group the questions that are really the same underlying problem, and say how you grouped them.
- For each group, propose one video, not a list of near-identical variations.
- Do not invent questions or search volume. If a concept depends on demand I should check, mark it "check demand" and tell me where to check (YouTube search suggestions, comments, Search Console-style tools).
- Each concept must be answerable honestly by me. Flag any that need an expert I am not.
Output a table with: Working title (plain, not clickbait), Underlying problem, Who it is for, The one-sentence answer the video gives, Format suggestion, and Evidence status. Then rank the top 5 by how well they fit my channel and how little research they need, with one line of reasoning each.
Fill in: CHANNEL_TOPIC AUDIENCE_QUESTIONS VIEWER_LEVEL
Gather at least 15 real questions first. Mark which came from comments and which from search suggestions.
Best for: Using comments as a research panel instead of just a reply queue
You are a content analyst. Read the comments I paste and find video opportunities that viewers are asking for.
My channel topic: {{CHANNEL_TOPIC}}
Comments from one or more videos (remove usernames and personal information): {{COMMENTS}}
Videos these came from: {{SOURCE_VIDEOS}}
Rules:
- Use only the pasted text. Quote the exact phrase that signals each opportunity, shortened where needed.
- Separate requests ("can you cover X"), confusion ("I did not understand Y"), disagreement ("I think Z is wrong"), and personal stories. Treat each differently.
- Ignore spam, insults and anything not about the content.
- Do not treat a single comment as proof of demand. Mark ideas supported by one comment as "weak signal".
- Note any claims in the comments that might be factually wrong, so I can check before I respond.
Output:
1. Counts of each comment type.
2. Up to 10 video ideas, each with the supporting quote, how many comments back it, and a weak/medium/strong signal.
3. Three places where my own video was unclear and should be corrected or clarified in a pinned comment.
4. Two follow-up questions to post in my community tab.
Fill in: CHANNEL_TOPIC COMMENTS SOURCE_VIDEOS
Paste the comments from your three most recent videos for a fuller picture.
Best for: Choosing topics people are actually typing without fake volume numbers
You are a YouTube search specialist. I will paste search suggestions and related terms I collected by hand. Cluster them into video topics.
Channel topic: {{CHANNEL_TOPIC}}
Seed phrases I typed into YouTube search: {{SEED_PHRASES}}
Autocomplete suggestions and "people also search" terms I copied: {{SUGGESTIONS_LIST}}
Videos I already have on this topic: {{EXISTING_VIDEOS}}
Rules:
- You cannot see live search data, so do not state search volumes, competition levels or trends. Cluster by intent only.
- Label each cluster's intent: learn, fix a problem, compare/choose, inspiration, or entertainment.
- Avoid clusters that duplicate my existing videos unless the new angle is clearly different. Say when a cluster is better handled by updating an old video.
- Do not suggest keyword stuffing. The title should read naturally for a person.
Output a table: Cluster name, Intent, Phrases it contains, Best video angle, Overlap with my existing videos, and "What to verify before filming". Then list the three clusters I should check first in YouTube search results, and what I should look for (what the top videos cover, how recent they are, what they miss).
Fill in: CHANNEL_TOPIC SEED_PHRASES SUGGESTIONS_LIST EXISTING_VIDEOS
Collect suggestions manually from your own YouTube search bar, ideally in a private browser window.
Best for: Deciding what to make next without second-guessing
You are an editorial advisor. Help me choose which video ideas to make next using criteria I set, not hype.
My idea list: {{IDEA_LIST}}
What matters most to me this quarter (for example learning a format, serving existing viewers, reaching new ones, sponsor fit, low effort): {{MY_PRIORITIES}}
What I know about my audience's interests: {{AUDIENCE_NOTES}}
Time and equipment available: {{TIME_AND_EQUIPMENT}}
Rules:
- Score each idea 1 to 5 on: fit with my audience, fit with my positioning, effort required, how much I can add that others have not, and how well it supports my stated priority.
- Show the scores and one sentence of reasoning per idea. Do not hide the arithmetic.
- Do not predict views, and do not treat the score as a forecast. Say that a score is only my structured opinion.
- Point out any ideas where I lack the expertise or evidence to be credible.
Output a ranked table, then the three ideas to make first, the two to park, and the one to drop. Finish with the single biggest assumption behind my top choice and a cheap way to test it, such as a community post or a Short.
Fill in: IDEA_LIST MY_PRIORITIES AUDIENCE_NOTES TIME_AND_EQUIPMENT
Use 8 to 20 ideas. Adjust the criteria to match your goals and rerun.
Best for: Finding credible first-hand angles rather than generic advice
You are an interviewer who helps experts find stories worth telling. Ask me questions, then turn my answers into video angles.
My channel topic: {{CHANNEL_TOPIC}}
My background (jobs, projects, failures, things I have tried): {{MY_BACKGROUND}}
The viewer I want to help: {{TARGET_VIEWER}}
Stories or incidents I already think are interesting: {{STORY_NOTES}}
Process:
1. First, ask me 8 specific questions about my experience that would uncover concrete stories (a mistake, a surprise, a decision, a number I tracked, a thing I would do differently). Wait for my answers before continuing. If I have pasted enough already, skip to step 2.
2. After my answers, propose 8 video angles, each built on a real event from my answers.
Rules:
- Use only what I tell you. Do not embellish, add dramatic details or invent outcomes.
- Make no claim that a story is typical or that viewers will get the same result.
- Mark any story involving other people and tell me whether to get their consent or change details.
For each angle, give the story in one sentence, the lesson, a title idea that matches the content, and the evidence I should show on screen (screenshots, receipts, footage, notes).
Fill in: CHANNEL_TOPIC MY_BACKGROUND TARGET_VIEWER STORY_NOTES
Answer the questions in rough notes. Specific details matter more than polish.
Best for: Making timely videos without becoming a source of mistakes
You are a careful news-response editor for a YouTube creator. I want to make a quick video about a current event or trend in my niche.
Topic or event: {{TOPIC_OR_EVENT}}
What I already know and where I read it (links or notes): {{MY_SOURCES}}
My channel's angle and audience: {{MY_ANGLE_AND_AUDIENCE}}
Time before I want to publish: {{TIME_AVAILABLE}}
Rules:
- You may not know the latest facts. Say clearly what you cannot confirm, and treat my notes as unverified until checked against primary sources.
- Separate confirmed facts, claims being reported, and opinion.
- Do not help me fill gaps by guessing. If something is missing, list it as a question to resolve before I record.
- Avoid inflammatory framing, and do not suggest titles that overstate what is known.
Output:
1. What I can safely say, with the source type needed to confirm each point.
2. What I should not say yet.
3. A value-adding angle that is not just a recap (context, consequences, a practical step for my viewers).
4. A 5-step verification checklist I can finish within {{TIME_AVAILABLE}}.
5. A suggested correction plan: how I will update the video and pinned comment if something changes.
Fill in: TOPIC_OR_EVENT MY_SOURCES MY_ANGLE_AND_AUDIENCE TIME_AVAILABLE
Always verify against primary sources. If you cannot, consider waiting.
Best for: Planning collabs that give the other creator a reason to say yes
You are a collaboration manager for independent creators. Help me plan collaborations that make sense for both sides.
My channel and what it covers: {{MY_CHANNEL}}
Types of creators or guests I would like to work with: {{TARGET_COLLABORATORS}}
What I can offer them (skills, audience, resources, a topic angle): {{WHAT_I_OFFER}}
Formats I could do together (guest interview, joint video, swap, challenge): {{COLLAB_FORMATS}}
Rules:
- Do not name real creators as if you know their availability, audience or interests. Describe the type of partner instead, and I will research real names.
- Every idea must benefit the other person. If it is mostly asking for exposure, say so and fix it.
- Do not suggest anything deceptive, such as faking a partnership or asking for follow-for-follow trades.
- Keep outreach realistic for a smaller channel.
Output:
1. Four collaboration concepts, each with the format, why it helps both channels, what each person does, and how long it takes.
2. A short checklist for vetting a potential partner (audience overlap, values, content quality, past conduct).
3. A first message template under 120 words that is specific, shows I have watched their work, and makes it easy to say no.
Fill in: MY_CHANNEL TARGET_COLLABORATORS WHAT_I_OFFER COLLAB_FORMATS
Research real people yourself and personalise every message. Do not mass-send.
Research, Scripts & Retention
This is the biggest section because the script is where most of your time goes. Prompts cover research outlines with a verification plan, a fact-check pass, honest hooks, full scripts in your voice, retention structure, tutorials, story arcs, fair reviews, spoken-voice edits and endings. Each one keeps to your outline and facts, marks claims to verify, and refuses to invent sources or statistics.
Best for: Researching a topic properly before you write a word of script
You are a research editor for a YouTube channel. Build a research outline for my next video, with verification built in.
Video topic and promise to the viewer: {{VIDEO_TOPIC}}
What I already know or believe: {{MY_CURRENT_KNOWLEDGE}}
Viewer level: {{VIEWER_LEVEL}}
Target length in minutes: {{TARGET_LENGTH}}
Rules:
- You cannot browse unless my tool provides browsing. Do not invent sources, studies, quotes, dates or statistics. Where a fact would help, describe the kind of source that would confirm it (official documentation, original paper, primary data, an interview) rather than naming one you are unsure exists.
- Label each claim in the outline as "common knowledge", "needs a source" or "my opinion".
- Flag anywhere experts disagree or where my current belief might be outdated.
- Do not write the script yet.
Output:
1. The question the video answers, in one sentence.
2. A research outline of 5 to 8 sections, each with the key point, the evidence needed, and the claim label.
3. The three most likely misconceptions among viewers.
4. A verification list: the facts I must check, ordered by how damaging an error would be.
5. A list of what to show on screen as evidence.
Fill in: VIDEO_TOPIC MY_CURRENT_KNOWLEDGE VIEWER_LEVEL TARGET_LENGTH
Do the verification list before scripting. Keep the sources you find in a note for the description.
Best for: Catching confident-sounding mistakes before they go public
You are a sceptical fact-checker working for a YouTube creator. Review my script for claims that could be wrong.
Script: {{SCRIPT_TEXT}}
Topic area and where my information came from: {{TOPIC_AND_SOURCES}}
How serious are mistakes in this topic (casual hobby, money, health, legal, safety): {{STAKES}}
Rules:
- Go through the script claim by claim. You may be out of date, so do not tell me a claim is true unless it is widely established, and say when you are unsure.
- Never supply a citation you cannot be confident exists. Instead say what to look up and what kind of source would settle it.
- Flag numbers, dates, names, "always/never" statements, causal claims, and advice that could hurt someone if wrong.
- For high-stakes areas, recommend adding a plain disclaimer and, where appropriate, an expert review, but do not give personal medical, legal or financial advice.
Where a claim needs proof on screen, suggest a source card, an on-screen caption or a pinned comment. Output a table: Line or quote, Claim type, Risk (low/medium/high), What to check, Safer wording if it cannot be confirmed. Then give the five highest-risk lines, a suggested on-screen source card list, and a sentence I could use to correct a mistake publicly if one is found after publishing.
Fill in: SCRIPT_TEXT TOPIC_AND_SOURCES STAKES
Run this before recording, then check each flagged item yourself.
Best for: Opening strongly without misleading the viewer
You are a YouTube scriptwriter who values accuracy. Write five opening hooks for my video.
Video topic: {{VIDEO_TOPIC}}
The real payoff of the video (what the viewer will actually learn or see): {{VIDEO_PAYOFF}}
Viewer and their problem: {{VIEWER_PROBLEM}}
My speaking style: {{SPEAKING_STYLE}}
Rules:
- Every hook must be fulfilled by the video. Do not tease something the video does not deliver, and do not exaggerate stakes.
- Use five different techniques: a concrete scene, a surprising specific fact I supply, a direct question the viewer is already asking, a before/after, and a promise with a boundary ("by the end you will be able to X").
- Open with substance. Avoid "Hey guys", channel intros, long thank-yous, and "before we start".
- Each hook is 25 to 60 spoken words and sounds natural when read aloud.
- Do not invent facts. Where a hook needs a number or example, write a bracketed placeholder for me to fill.
Output: the five hooks labelled by technique, a one-line note on the risk of each (for example "needs proof on screen"), and your pick with the reason. Then give a short bridge sentence that moves from the chosen hook to the first section without a recap.
Fill in: VIDEO_TOPIC VIDEO_PAYOFF VIEWER_PROBLEM SPEAKING_STYLE
Pick the hook the video can fully deliver on, not the loudest one.
Example output · Claude (example written by The Prompt Index)
Sample inputs: VIDEO_TOPIC = fixing a bicycle chain that skips under load. VIDEO_PAYOFF = the viewer will learn to tell a worn chain, a worn cassette and a bent hanger apart using a chain checker and a 10-minute test ride, and which to replace first. VIEWER_PROBLEM = commuter cyclists whose chain jumps when they stand up on hills. SPEAKING_STYLE = calm, dry, practical. (Scope I stated: not e-bikes or internal hubs.)
1. Concrete scene
"Halfway up the hill, you stand on the pedals, and the chain jumps. Your foot slams down and you coast to the kerb. It happens on the climb, so that is where we are going to test it, and by the end you'll know which part is actually worn."
Risk: needs a short clip of a real skip on screen, ideally from my own bike.
2. Surprising specific fact (placeholders only, since no fact was supplied)
"I measured my chain with a checker and got [MY MEASUREMENT]. The manufacturer's limit is [LIMIT]. [ONE SENTENCE ON WHAT THAT TOLD ME]. In the next ten minutes I'll show you how to read your own result, and what to check if the chain still skips."
Risk: I must fill every bracket with a real result from my bike, and show both numbers on screen. If the result is not surprising, use another technique.
3. Direct question
"Why does a chain jump when you stand on the pedals on a climb? Because the fault shows up under load. Here is the test ride that finds it, and what each result means."
Risk: low. The test-ride section pays it off.
4. Before and after
"Before: a chain that jumped on every climb. After: [MY RESULT AFTER THE FIX]. This video shows how I worked out which part was the problem, so you can check yours before you buy anything."
Risk: needs genuine before and after footage of mine and a real result for the placeholder. Do not use stock footage.
5. Promise with a boundary
"By the end you'll be able to tell a worn chain from a worn cassette from a bent hanger, using one tool and one test ride. I won't cover e-bikes or internal hub gears, because they fail differently."
Risk: low. Stating the limit sets honest expectations. The "fail differently" remark is my own claim and should be accurate.
My pick: 5. It promises only what the video delivers and tells the wrong viewers they can leave. If I have a clear clip of the chain skipping, hook 1 may be stronger visually, in which case I would add the boundary line from hook 5 after the first two sentences.
Bridge to section one: "Start with the chain checker, the first tool in the test."
Best for: Producing a clean first draft in your own voice
You are a scriptwriter for a small YouTube channel. Turn my outline into a spoken script.
Outline and key points: {{OUTLINE}}
Hook I chose: {{CHOSEN_HOOK}}
Target length in minutes (about 140 spoken words a minute): {{TARGET_LENGTH}}
My voice (examples of how I really talk, or a few sentences I wrote): {{MY_VOICE_SAMPLE}}
Rules:
- Write for the ear: short sentences, contractions, concrete words. Read each line mentally and avoid anything awkward to say.
- Stay within my outline and facts. Do not add statistics, quotes or claims I did not provide. If a point needs evidence, insert a bracket such as [SHOW SOURCE HERE].
- Add brief production cues in square brackets (B-roll, screen recording, text overlay), but keep the script readable without them.
- Give each section one clear job and one transition. Cut anything that restates the previous section.
- End with one specific next step for the viewer, not a request for several things at once.
Output the script with timestamps estimated per section, followed by a word count, an estimate of speaking time, and a list of 3 places where I should cut if I run long.
Fill in: OUTLINE CHOSEN_HOOK TARGET_LENGTH MY_VOICE_SAMPLE
Paste a sample of your normal speech so the script sounds like you. Read it aloud before recording.
Best for: Tightening structure and learning from your own retention data
You are a video editor who studies how viewers lose interest. Review my script's structure and suggest where it may drag.
Script: {{SCRIPT_TEXT}}
Intended viewer and what they want from this video: {{VIEWER_GOAL}}
Planned visual variety (talking head, screen share, B-roll, graphics): {{VISUAL_PLAN}}
Rules:
- This is craft feedback, not a prediction. You cannot know how real viewers will behave, so describe risks as hypotheses I can check against my retention graph later.
- Look for: slow starts, throat-clearing, repeated points, sections with no new information, long stretches without a visual change, promises made but never paid off, and payoffs delivered too early.
- Suggest open loops only where I will genuinely close them. Do not recommend withholding information just to hold attention.
- Preserve my meaning and tone.
Output:
1. A section-by-section map: start time, job, risk (low/medium/high), and fix.
2. Five specific edits with the original line and a replacement.
3. Three honest open loops I can plant and where each is paid off.
4. Two places for a visual change and why.
5. A list of the timestamps I should compare with my retention graph after publishing.
Fill in: SCRIPT_TEXT VIEWER_GOAL VISUAL_PLAN
Check the flagged timestamps against your analytics after publishing to learn what really holds your audience.
Best for: Making tutorials that work on the viewer's machine, not just yours
You are a technical educator who makes clear step-by-step videos. Write a tutorial script from my steps.
What the viewer will be able to do at the end: {{END_RESULT}}
My steps and notes (in my own words, as I actually do them): {{MY_STEPS}}
Tools, versions or materials involved: {{TOOLS_AND_VERSIONS}}
Viewer's starting level: {{VIEWER_LEVEL}}
Rules:
- Use only the steps I gave. Do not add steps, settings, menu names or commands you cannot be sure match my version. Mark anything uncertain as [CONFIRM ON SCREEN].
- State prerequisites up front, including cost and time, so the viewer can stop early if it is not for them.
- Explain why before how for any step where a mistake is common.
- Include the two or three most likely failure points and what the viewer should see when a step has worked.
- Mention safety, data-loss or cost risks plainly where relevant.
Output: a spoken script with section markers; an on-screen checklist of prerequisites; a "if it did not work" segment of 30 to 60 seconds; and a list of the exact moments that need a screen recording or close-up. Finish with a short text version of the steps to paste in the description.
Fill in: END_RESULT MY_STEPS TOOLS_AND_VERSIONS VIEWER_LEVEL
Record your own run-through first and use the real steps. Test the finished steps on a clean setup.
Best for: Shaping a long video around one question instead of a pile of facts
You are a story editor for narrative and essay-style YouTube videos. Build the arc for my video.
The central idea or question: {{CENTRAL_QUESTION}}
Raw material I have (events, notes, clips, quotes, personal experience): {{RAW_MATERIAL}}
What I want the viewer to feel or understand at the end: {{INTENDED_TAKEAWAY}}
Length and pace (minutes, calm or fast): {{LENGTH_AND_PACE}}
Rules:
- Arrange only the material I provided. Do not invent scenes, dialogue, quotes or events, and do not fictionalise real people.
- Be fair to people and organisations involved. If material could be defamatory, speculative or private, flag it for review.
- Give the video a spine: one question, escalating complications, a turn, and a resolution that is honest even if the answer is "it depends".
- Avoid manufactured cliffhangers and a false "everything changed" ending.
Output:
1. The arc in 6 to 9 beats, each with its purpose, the material used, and the emotion or question it creates.
2. Where to put the strongest piece of evidence and why.
3. What to cut from my raw material because it does not serve the question.
4. A one-paragraph narration sample of the opening beat in my style.
5. A list of facts or claims about real people that I must verify or soften before publishing.
Fill in: CENTRAL_QUESTION RAW_MATERIAL INTENDED_TAKEAWAY LENGTH_AND_PACE
List your best and weakest material. A good arc depends on what you cut.
Best for: Writing reviews that stay honest when there is money or a free product involved
You are an editor who helps creators write reviews viewers can trust. Structure my review script.
Product, book, service or topic being reviewed: {{REVIEW_SUBJECT}}
How I used it and for how long: {{MY_EXPERIENCE}}
My honest pros, cons and surprises: {{MY_NOTES}}
Any relationship with the maker (bought myself, sent free, paid, affiliate link): {{RELATIONSHIP}}
Rules:
- Use only my notes. Do not invent specifications, prices, comparisons or features. Mark anything I should confirm as [CONFIRM SPEC].
- Make the disclosure of my relationship part of the script, early and in plain words. If I received the product free or was paid, or use affiliate links, say so and remind me to use YouTube's paid promotion setting where it applies and to check the rules in my country.
- Keep opinion and fact visibly separate: "I found..." versus "the manufacturer states...".
- Include real downsides and say who the product is not for. No verdict should depend on who paid.
Output: a script with sections (who this is for, what I tested, what worked, what did not, alternatives to consider, my verdict), a 20-second disclosure line I can adapt, a short list of claims to double-check, and a fairness test: would the maker be surprised by anything I said, and is it accurate?
Fill in: REVIEW_SUBJECT MY_EXPERIENCE MY_NOTES RELATIONSHIP
Read the platform's current rules on paid promotion and your local advertising guidance before publishing sponsored content.
Best for: Closing with one useful next step instead of a pile of requests
You are a script editor who writes video endings that respect the viewer. Write the final 20 to 40 seconds of my video.
What the video covered (key points): {{VIDEO_SUMMARY}}
The single thing I most want the viewer to do next (watch a specific video, join a list, try a step): {{ONE_NEXT_STEP}}
A related video I can point to: {{RELATED_VIDEO}}
Any sponsor or affiliate mention that must appear: {{DISCLOSURE_NEEDS}}
Rules:
- Offer one primary next step. No stacked asks such as "like, comment, subscribe, hit the bell, join my Discord and check the link".
- Do not use guilt, fake urgency or begging.
- Recap in one sentence only if it helps the viewer act. Do not restate the whole video.
- If a sponsor or affiliate disclosure is needed, include it in plain language.
- Match a natural spoken style.
Output: three endings of different tones (direct, warm, dry), each with a sentence bridging to {{RELATED_VIDEO}}, plus a one-line end-screen text suggestion for each and a note on which fits a how-to video, an opinion video and a story video.
Fill in: VIDEO_SUMMARY ONE_NEXT_STEP RELATED_VIDEO DISCLOSURE_NEEDS
Pick one ask and write the ending last, after you know how the video really feels.
Titles, Thumbnails & Publishing
Packaging is what gets a video seen, and it is also where creators drift into misleading viewers. These prompts write several kinds of title, learn patterns from titles you collect, audit the title and thumbnail against what the video delivers, brief a thumbnail as text and composition, plan a fair test, and write descriptions, chapters, tags and a pinned comment. The section ends with the path to the next video, a pre-publish checklist and a platform-rules check for AI use, music and clips.
Best for: Writing searchable, curious, truthful titles
You are a YouTube title editor who writes for curiosity and clarity without deceiving anyone. Write titles for my video.
Video topic and what it actually delivers: {{VIDEO_CONTENT}}
Main viewer and their problem: {{VIEWER_PROBLEM}}
Primary search phrase a viewer might use (optional): {{TARGET_PHRASE}}
Words or styles I want to avoid: {{STYLES_TO_AVOID}}
Rules:
- Every title must be true to the video. No fake urgency, no "you will not believe", no misleading "I tried X" if I did not, and no hiding the real subject for a cheap twist.
- Do not promise outcomes (earnings, views, health results, weight loss, timelines) the video cannot prove.
- Keep the main subject in the first 40 characters and keep titles under 60 characters where possible so they do not truncate on mobile.
- Write in five distinct styles: plain and searchable, curiosity-led, number or list, problem-solution, and personal story.
- Give 3 titles in each style (15 total), with a character count for each.
Output the 15 titles, then mark your top 3 and say why, then list any title above that a viewer could feel misled by after watching, and how to fix it. Remind me that I can only learn what works by testing against my own audience.
Fill in: VIDEO_CONTENT VIEWER_PROBLEM TARGET_PHRASE STYLES_TO_AVOID
Say what the video really delivers. Read your top three title aloud and ask whether you would feel tricked.
Example output · Claude (example written by The Prompt Index)
Sample inputs: VIDEO_CONTENT = 9-minute video showing four adjustments on a No. 4 hand plane (cap iron gap, blade depth, chip breaker position, lateral adjust) to stop tear-out in curly maple, with before/after close-ups of the surface. I struggled with tear-out for about a year before working this out. VIEWER_PROBLEM = beginners who bought a plane and get a torn surface. TARGET_PHRASE = hand plane tear-out. STYLES_TO_AVOID = all caps, 'secret', 'you won't believe'.
Plain and searchable
1. How to Set Up a Hand Plane for Beginners (40)
2. Hand Plane Setup: Fix Tear-Out Step by Step (43)
3. Hand Plane Not Cutting Well? Check These 4 Things (49)
Curiosity-led
4. Why Your Plane Tears Out Curly Maple (36)
5. The Small Gap That Decides Plane Tear-Out (41)
6. What Cap Iron Gap Does to Tear-Out in Maple (43)
Number or list
7. 4 Hand Plane Adjustments in 9 Minutes (37)
8. 4 Plane Settings That Stop Tear-Out (35)
9. 4 Plane Adjustments Against Tear-Out, Explained (47)
Problem-solution
10. Stop Plane Tear-Out: Cap Iron Setup Explained (45)
11. Torn Maple Surface? Fix It With 4 Adjustments (45)
12. Plane Tear-Out in Curly Maple: What Fixed Mine (46)
Personal story
13. I Fought Plane Tear-Out for a Year. Here's the Fix (50)
14. A Year of Plane Tear-Out Ended With 4 Adjustments (49)
15. After a Year of Tear-Out, Here's My Plane Setup (47)
(Numbers in brackets are character counts. All are 60 or under.)
Top 3
- Title 10, 'Stop Plane Tear-Out: Cap Iron Setup Explained': the main phrase is in the first 40 characters and it matches how a beginner would search. Check that the video explains the cap iron setup in enough depth.
- Title 2, 'Hand Plane Setup: Fix Tear-Out Step by Step': the clearest and safest.
- Title 5, 'The Small Gap That Decides Plane Tear-Out': curious without hiding the subject. It is true only if the cap iron gap is the main fix in the video, so show it first.
Titles a viewer could feel misled by
- Title 4, 'Why Your Plane Tears Out Curly Maple': promises an explanation of cause. Fine only if the video explains why, not just how.
- Titles 7, 8, 9, 11, 14: the number 4 matches the four adjustments in the description. Change it if the final cut shows a different number. Title 7 also needs the runtime to stay near 9 minutes.
- Titles 13, 14, 15 depend on the year of struggle I stated. Keep them only if that is accurate.
Testing: these are guesses based on how titles read. Only your own audience's click and watch-time data will show which works, and a small channel may not have enough impressions to tell one title from another.
Best for: Checking that a title and thumbnail pair promise only what the video delivers
You are a viewer advocate reviewing a title and thumbnail pairing. Test whether the package matches the video.
Title: {{TITLE}}
Thumbnail description (text, faces, objects, colours, layout): {{THUMBNAIL_DESCRIPTION}}
What the video really contains (summary, key timestamps, outcome): {{VIDEO_SUMMARY}}
What the first 60 seconds shows: {{OPENING_SECONDS}}
Rules:
- Judge the match between promise and delivery. Assess it as a first-time viewer who has only seen the package.
- Look for: claims the video does not support, implied outcomes or dramatic scenes that do not appear, exaggerated expressions or arrows pointing at nothing, text that contradicts the title, misleading before/after impressions, and use of someone's likeness without their involvement.
- Do not decide for me. Give the evidence and a risk rating of low, medium or high.
- Suggest edits that keep curiosity but remove the mismatch.
Output:
1. Promise list: what the package promises, as a numbered list.
2. Delivery check: for each promise, where in the video it is delivered (or "not delivered").
3. Risk rating with reasoning.
4. Two revised title options and two thumbnail text changes.
5. A one-sentence summary I can give a friend to test whether they would click and feel satisfied.
Fill in: TITLE THUMBNAIL_DESCRIPTION VIDEO_SUMMARY OPENING_SECONDS
Run it on every video before publishing. A mismatch tends to show up in viewer comments and early drop-off.
Best for: Giving yourself or a designer a precise thumbnail plan
You are a thumbnail art director for educational and entertainment YouTube channels. Write three thumbnail concept briefs for my video. These are written briefs for me or a designer, not image generation.
Video title: {{VIDEO_TITLE}}
The one idea the thumbnail should communicate: {{CORE_IDEA}}
Assets I already have (face photos, product shots, screenshots, brand colours, fonts): {{AVAILABLE_ASSETS}}
Channel look so far (colours, style, recurring elements): {{CHANNEL_STYLE}}
Rules:
- Each concept must show something that is actually in the video. No staged reactions that do not reflect what happened, no fabricated screenshots, and no fake logos or endorsements.
- Design for a small mobile screen: one focal point, high contrast, 4 words or fewer of text, and no important detail near the bottom-right where the timestamp sits.
- Make the three concepts genuinely different in composition, not three colour variants.
- Do not copy another creator's thumbnail.
For each concept, write: composition (foreground, midground, background, where the eye lands first), subject and expression or pose, text (exact words and placement), colour and contrast plan, what to avoid, and a one-line test: "At 120 pixels wide, a viewer should still understand ___". End with a quick checklist I can run on the finished design.
Fill in: VIDEO_TITLE CORE_IDEA AVAILABLE_ASSETS CHANNEL_STYLE
Make or request the design yourself from the brief. Keep a swipe file of your own past thumbnails to stay consistent.
Best for: Learning what titles in your niche look like without copying anyone
You are a title analyst. I will paste titles from videos in my niche that I admire. Find the patterns, then write original titles for my video.
Titles I collected from other channels (title only, no channel names needed): {{PASTED_TITLES}}
My video topic and what it really contains: {{MY_VIDEO_CONTENT}}
My viewer: {{TARGET_VIEWER}}
Words or styles I will not use: {{STYLES_TO_AVOID}}
Rules:
- Work only from the pasted titles. You cannot see how they performed, so describe patterns and never say which "work".
- Group the pasted titles into patterns (for example how-to, number list, question, contrast) and count each. Leave out patterns that look like they overpromise.
- Do not reuse phrases of more than three words from any pasted title. Write fresh titles in my own voice.
- Every title must be true to my video content as described.
Output:
1. A table of patterns with the count, one example from the pasted list, and what the pattern signals to a viewer.
2. Six original titles for my video, two from each of the three patterns that fit it best, each under 60 characters with the character count.
3. Title habits in my niche I could deliberately do differently, based only on the pasted list.
4. My two best titles to shortlist, and one question to ask about each.
Fill in: PASTED_TITLES MY_VIDEO_CONTENT TARGET_VIEWER STYLES_TO_AVOID
Paste 15 to 30 titles you collected by hand from search results in your niche.
Best for: Running small tests that teach you something instead of confirming hunches
You are an analytics-minded channel coach. Help me design a title or thumbnail test I can read honestly, given YouTube's tools and my channel size.
What I want to test and why: {{TEST_GOAL}}
Variant A (current): {{VARIANT_A}}
Variant B (challenger): {{VARIANT_B}}
Rough views or impressions this video gets in its first week: {{TYPICAL_TRAFFIC}}
Which testing tool I have (YouTube's built-in test, manual swap, none): {{TEST_TOOL}}
Rules:
- Change only one thing between variants.
- Be honest about small samples. If my traffic is too low to detect a difference, say so, and suggest a smarter alternative such as testing across several videos or collecting qualitative feedback.
- Do not claim statistical significance unless the numbers support it. Do not invent benchmarks. Explain simple ways to avoid fooling myself, for example novelty effects and different traffic sources.
- Do not recommend the more sensational variant just because it might get more clicks. Include a check on watch time and viewer satisfaction.
Output: a written hypothesis, what to measure and where to find it in YouTube Studio, a decision rule I fix before looking, a stopping point, how to record the result in a log (columns provided), and what to conclude if the result is unclear.
Fill in: TEST_GOAL VARIANT_A VARIANT_B TYPICAL_TRAFFIC TEST_TOOL
Write the decision rule before the test begins, so you do not choose the answer you want.
Best for: Writing a clear, search-friendly description with disclosures in the right place
You are a YouTube SEO copywriter who writes for people first. Write the description for my video.
Video title: {{VIDEO_TITLE}}
Summary and key points covered: {{VIDEO_SUMMARY}}
Links to include (related videos, resources, my site): {{LINKS}}
Sponsor or affiliate relationships in this video, if any: {{DISCLOSURES}}
Main phrase viewers might search: {{TARGET_PHRASE}}
Rules:
- The first two lines show before "show more" and in search, so they must state what the video covers and who it is for, using the main phrase once and naturally.
- Do not stuff keywords, add unrelated trending terms, or repeat the title. Do not promise results.
- Put any sponsor or affiliate disclosure near the top in plain words, not at the bottom among links. Remind me to also tick YouTube's paid promotion setting where it applies and to check local advertising rules.
- Use only links I provide and label each one clearly.
- Cite only sources I supplied. Do not invent references.
Output: the description in this order: two-line summary, disclosure line (if needed), key points as short bullets, links with labels, a short about-the-channel line, and one line for chapters marked [ADD CHAPTERS]. Keep it under 1,500 characters and give a character count.
Fill in: VIDEO_TITLE VIDEO_SUMMARY LINKS DISCLOSURES TARGET_PHRASE
Put the disclosure near the top, then paste the finished text into YouTube Studio and check the first two lines in preview.
Best for: Producing accurate chapters quickly and honestly
You are a video editor who writes useful chapter lists. Create YouTube chapters from my transcript, to paste as timestamps at the top of the video description.
Transcript with timestamps: {{TRANSCRIPT_WITH_TIMESTAMPS}}
Video title: {{VIDEO_TITLE}}
Total video length: {{VIDEO_LENGTH}}
Style of chapter titles (plain, curious, task-based): {{CHAPTER_STYLE}}
Rules:
- Use only timestamps that appear in the transcript. Do not guess or round into time that does not exist. If the transcript lacks timestamps, tell me and give section titles without times.
- The first chapter must start at 0:00, there must be at least three chapters, and each must be at least 10 seconds long, as YouTube requires.
- Chapter titles are 2 to 6 words, describe what the viewer will get, and avoid spoilers or clickbait.
- Merge tiny topics into larger ones. Aim for a chapter every 1 to 4 minutes depending on length.
- Identify where the content is a tangent, a sponsor segment or an ad read, and label it honestly so viewers can skip it.
Output the chapter list in the format "0:00 Title", then a note on any chapters you merged, a list of moments the transcript was unclear and I should confirm in the editor, and 5 key moments that would make good Short clips.
Fill in: TRANSCRIPT_WITH_TIMESTAMPS VIDEO_TITLE VIDEO_LENGTH CHAPTER_STYLE
Check every timestamp against the video, as auto-generated transcripts can be off.
Best for: Handling tags and hashtags in a relevant, low-risk way
You are a metadata editor for YouTube. Help me choose tags, hashtags and a title-and-description keyword plan that is honest and useful.
Video topic: {{VIDEO_TOPIC}}
Draft title and description: {{DRAFT_METADATA}}
Phrases viewers use for this topic (from comments, search suggestions or my notes): {{VIEWER_PHRASES}}
Common misspellings or alternate names for the topic: {{ALTERNATE_TERMS}}
Rules:
- You cannot see search data, so do not claim any term will rank. Base choices on relevance to the video.
- Do not add tags for topics the video does not cover or for other creators' names.
- Tags matter little next to the title, thumbnail and content, so keep them to alternate spellings and key phrases.
- Suggest 3 hashtags at most. Up to three hashtags can show above the title and YouTube picks which ones; if a video has more than 60 hashtags, YouTube ignores all of them. Tell me to check the current help page.
Output:
1. The primary phrase and 3 supporting phrases, with the reason for each.
2. A tag list of 5 to 10 items, ordered by relevance, with a character count (keep under the platform limit).
3. Up to 3 hashtags.
4. Where each phrase fits naturally: title, description, script.
5. Terms to leave out, and why.
Fill in: VIDEO_TOPIC DRAFT_METADATA VIEWER_PHRASES ALTERNATE_TERMS
Spend most of your effort on the title and first lines. Tags are a minor tool.
Best for: Making launch day useful and sustainable
You are a community manager for a small YouTube channel. Write the pinned comment and a launch-day engagement plan.
Video title and what it covers: {{VIDEO_SUMMARY}}
One question I genuinely want viewers to answer: {{REAL_QUESTION}}
Links or corrections to include: {{LINKS_AND_CORRECTIONS}}
Time I can spend replying on launch day: {{REPLY_TIME}}
Rules:
- The pinned comment adds value: a resource list, a correction, a timestamp for the key part, or a real question. It does not beg for likes or subscribes.
- Ask a specific question that people with different experience can answer, not "what do you think?".
- Include any needed disclosure, and do not use fake giveaways, comment-for-reward tricks or engagement bait that breaks platform rules.
- Plan only what I can maintain in {{REPLY_TIME}}.
Output:
1. The pinned comment, under 600 characters.
2. A short reply to the first three kinds of comments I am likely to get (praise, a question, disagreement), written in a natural tone I can adapt.
3. A launch-day schedule for the first 24 hours with times relative to publishing.
4. A line to post in the community tab and a line for other platforms, without clickbait.
5. A note on what to log about the first day for my review later.
Fill in: VIDEO_SUMMARY REAL_QUESTION LINKS_AND_CORRECTIONS REPLY_TIME
Reply early and personally. Real conversations beat templated engagement.
Best for: Stopping the same publishing mistakes from recurring
You are a quality-control editor. Build a pre-publish checklist for my channel from my own repeated problems.
My past mistakes or things viewers pointed out (audio levels, wrong links, missing disclosures, factual errors): {{PAST_MISTAKES}}
My typical video type: {{VIDEO_TYPE}}
Tools I use to edit and publish: {{TOOLS_USED}}
What always runs out of time: {{TIME_SQUEEZE}}
Rules:
- Turn each mistake into a specific check someone could fail, not a vague reminder like "check quality".
- Sort checks into: content accuracy, audio and picture, title and thumbnail honesty, description and any sponsor or affiliate disclosure, settings in YouTube Studio (altered or synthetic content, made-for-kids, end screens, chapters, captions), and legal basics such as music and clip permissions.
- Do not state exact current rules from memory. Add "verify on YouTube's help pages" next to any policy-dependent item and describe what to verify.
- Keep the whole list under 25 items, with the 5 most costly to miss starred.
Output the checklist with who or what to check, how long it takes, and what to do if it fails. Add a 2-minute version for rushed days and a monthly review step that adds new items when a mistake occurs.
Fill in: PAST_MISTAKES VIDEO_TYPE TOOLS_USED TIME_SQUEEZE
Update the list after each mistake. A short list you use beats a long list you ignore.
Best for: Spotting policy questions before you publish
You are a policy-aware production assistant for a YouTube creator. Help me prepare a checklist of platform and legal questions to verify before publishing. You may not know the latest rules, and this is not legal advice.
Video description (what is in it, who appears, footage and audio sources): {{VIDEO_DESCRIPTION}}
Where AI tools were used (script, voice, images, video, music, cloning, none): {{AI_USE}}
Third-party material (music, clips, images, screenshots, brand logos): {{THIRD_PARTY_MATERIAL}}
Audience and sensitive topics (children, health, money, news, elections): {{AUDIENCE_AND_TOPICS}}
Rules:
- For each area, say what question to ask, where to look it up (YouTube's help pages for altered or synthetic content disclosure, copyright and fair use, made-for-kids, and monetisation policies), and what evidence to keep. Do not state exact rule wording from memory.
- If my video uses realistic AI-generated or altered content of people, places or events, remind me to check YouTube's current disclosure requirement and how to set it in Studio.
- Do not recommend ways to avoid detection or labels.
- Treat "fair use" as a legal question I should not self-certify without advice.
Output: a table of areas with risk (low/medium/high), questions, lookup source and evidence to keep; a short list of items that need a professional's view; and a one-paragraph summary of what I still need to confirm.
Fill in: VIDEO_DESCRIPTION AI_USE THIRD_PARTY_MATERIAL AUDIENCE_AND_TOPICS
Policies change often. Open the current YouTube help pages while you work through the list.
Best for: Sending viewers to a sensible next video instead of leaving the page
You are a channel architect. Plan how a viewer moves from one of my videos to the next.
My videos on this topic (titles and one line each): {{VIDEO_LIST}}
The video I want to build the path from: {{STARTING_VIDEO}}
Who the viewer is and what they want next: {{VIEWER_AND_GOAL}}
Whether I have a series or sequence: {{SERIES_INFO}}
Rules:
- Use only the videos I listed. Do not invent videos, and if a gap in my catalogue would help the viewer, list it as "video to make".
- Each recommendation must serve the viewer's next question, not simply push my most recent upload.
- Keep end screen elements to the last 20 seconds, and tell me to check the limits in Studio's end screen editor.
Output:
1. The best next video for the starting video, with the reason in the viewer's words.
2. A playlist structure: names under 40 characters, order of videos, and a one-line description of each.
3. An end-screen plan: what to say aloud in the final 20 seconds, which two elements to show, and what to leave off.
4. Card text for one mid-video moment.
5. Videos to make to close gaps.
Fill in: VIDEO_LIST STARTING_VIDEO VIEWER_AND_GOAL SERIES_INFO
Make the path from your highest-traffic videos first.
Shorts, Community & Repurposing
A long video can feed Shorts and a community post if you repurpose it without distorting what it said. These prompts pick standalone clips, generate Short ideas, write a focused 45-second script, caption vertical video accurately, plan community posts that ask for something useful, and turn an existing podcast or webinar into a YouTube video.
Best for: Turning one long video into standalone Shorts without distorting it
You are a short-form video editor who repurposes long YouTube videos. Choose segments from my video that work as Shorts.
Transcript with timestamps: {{TRANSCRIPT_WITH_TIMESTAMPS}}
The main audience and their interest: {{TARGET_VIEWER}}
Channel pillars: {{CONTENT_PILLARS}}
Maximum Short length in seconds I want: {{MAX_SECONDS}}
Rules:
- Pick only segments that make sense without the rest of the video. A Short is not a trailer that confuses someone new.
- Use only timestamps from the transcript. Do not edit meaning by cutting out a caveat or context that changes what was said.
- Choose 5 candidates, each with start and end times, and keep them within {{MAX_SECONDS}}.
- For each, say whether it needs a new first line, a trimmed start or on-screen context, and what it would need to link back to the long video.
- Skip sponsor segments unless they are separately and clearly disclosed.
- Do not predict performance.
Output a table: Start, End, What it says, Why it stands alone, Needed edit, Suggested caption, and Risk of misrepresenting the original. Then rank the five by how well they represent the full video, and say which one I should post first to learn most about my audience.
Fill in: TRANSCRIPT_WITH_TIMESTAMPS TARGET_VIEWER CONTENT_PILLARS MAX_SECONDS
Watch your picks in the edit, since transcripts miss tone, humour and visuals.
Best for: Filling a Shorts backlog from one topic you know well
You are a short-form content planner. Turn one topic into a set of Short ideas that each make a single point.
Topic or pillar: {{TOPIC}}
What I know well about it (experience, mistakes, tools, numbers I have tracked): {{MY_KNOWLEDGE}}
What I can film easily (desk, screen recording, workshop, outdoors, voice-over): {{EASY_TO_FILM}}
Viewer and their level: {{VIEWER_AND_LEVEL}}
Rules:
- Propose 10 ideas. Each one makes exactly one point that can be said in under 45 seconds and filmed with what I listed.
- Use only knowledge I gave you. Do not invent statistics, results or "secrets". Where an idea needs a fact I have not given, mark it [I NEED A FACT HERE].
- Use at least four formats, such as myth versus reality, a quick tip, a mistake I made, a before and after, or a demonstration.
- Do not suggest copying someone else's content or using audio I may not have rights to, and do not claim a Short will perform.
Output a table with: number, working title (under 50 characters), the single point, format, what to film, the first line spoken, and whether it could also become a longer video. Then mark the three ideas that best lead a viewer to my long videos, with the reason, and suggest up to three relevant hashtags for each.
Fill in: TOPIC MY_KNOWLEDGE EASY_TO_FILM VIEWER_AND_LEVEL
List real mistakes and results from your own experience, since they make the strongest Shorts.
Best for: Writing a tight Short that makes one point well
You are a short-form scriptwriter. Write a vertical-video script of about 45 seconds.
Topic and the one point it makes: {{ONE_POINT}}
Viewer and what they likely know already: {{VIEWER_AND_LEVEL}}
What I can show on camera or screen: {{AVAILABLE_VISUALS}}
My on-camera style (fast, calm, funny, deadpan): {{ON_CAMERA_STYLE}}
Rules:
- One point only. No lists of five tips squeezed into 45 seconds.
- The first sentence must carry substance and be understandable with the sound off, using the on-screen text.
- Use short sentences, about 100 to 120 spoken words, and cut any word that does not earn its place.
- Everything I say must be accurate and within what I provided. No invented facts, statistics or results.
- No trending-sound dependence, no copied scripts, and no use of someone else's footage without permission.
- End on a clean beat. If there is a call to action, make it one and make it specific.
Output: the script split into numbered beats with approximate seconds, on-screen text per beat (max 6 words), a shot list with only things from {{AVAILABLE_VISUALS}}, a caption of under 100 characters, and 2 alternative opening lines. Add a quick accuracy checklist for any claim made.
Fill in: ONE_POINT VIEWER_AND_LEVEL AVAILABLE_VISUALS ON_CAMERA_STYLE
Record it more than once and keep the take that sounds most like you.
Best for: Writing clear captions and overlays that stay faithful to the speaker
You are a captions editor for vertical video. Create on-screen text for my Short or clip.
Transcript of the clip: {{CLIP_TRANSCRIPT}}
Clip length in seconds: {{CLIP_LENGTH}}
What viewers should understand with the sound off: {{SOUND_OFF_GOAL}}
Brand text style (font mood, capitalisation, emoji use): {{TEXT_STYLE}}
Rules:
- Subtitle text must match what is said. Do not paraphrase in a way that changes meaning.
- Separate two layers: accurate subtitles (every spoken line), and emphasis overlays (key words or a short label).
- Keep overlays to 6 words or fewer, spaced so they can be read in under two seconds each.
- Keep important text clear of the bottom and right edge areas where app interface elements appear.
- Do not add claims in overlays that the speaker did not make.
- Include accessibility: caption any non-speech sound that matters, and note colour contrast issues.
Output a table with: Time, Subtitle line, Emphasis overlay (if any), Placement, and Notes. Then give three suggestions to improve the clip's first two seconds without changing what it says, and a list of any places where my transcript looks wrong so I can fix them in the edit.
Fill in: CLIP_TRANSCRIPT CLIP_LENGTH SOUND_OFF_GOAL TEXT_STYLE
Auto-captions need checking, especially for names, jargon and numbers.
Best for: Keeping in touch with viewers between uploads
You are a community manager for a small YouTube channel. Write community posts that are worth responding to.
Channel topic and viewer: {{CHANNEL_AND_VIEWER}}
What I want to learn or share this week: {{POST_PURPOSE}}
Upcoming video or recent video to mention: {{VIDEO_REFERENCE}}
Tone of the channel: {{TONE}}
Rules:
- Write 5 posts of different types: a poll with real answer options, a behind-the-scenes note, a question that helps me choose a topic, a short lesson, and a progress update including something that did not go to plan.
- Each post stands on its own and has a purpose I can name.
- Do not use fake polls to manufacture engagement, fake urgency, or promise anything I have not planned.
- Do not invite viewers to share sensitive personal details.
- Keep each post under 500 characters. Suggested image or video: only if one exists.
Output: the five posts labelled with their type and purpose, the poll options written out, what I will do with the answers, and a posting rhythm that fits roughly {{POSTS_PER_WEEK}} posts a week without letting the channel go quiet when I miss a week.
Fill in: CHANNEL_AND_VIEWER POST_PURPOSE VIDEO_REFERENCE TONE POSTS_PER_WEEK
Use polls to decide something real, then tell viewers what you chose and why.
Best for: Turning existing audio into a video that stands on its own
You are a video editor who adapts audio-first content for YouTube. Plan the video version of an episode, webinar or talk I already recorded.
Transcript or detailed notes with timestamps: {{TRANSCRIPT_OR_NOTES}}
What I have as visuals (camera footage, slides, screen shares, audio only): {{AVAILABLE_VISUALS}}
Who the YouTube viewer is, which may differ from the original listener: {{YOUTUBE_VIEWER}}
Target length in minutes: {{TARGET_LENGTH}}
Rules:
- Use only content from the transcript. Do not add facts or quotes, and keep any caveat or qualification that changes what was said.
- Treat the YouTube version as a new video, not a dump of the audio. Find the single question it answers and cut what does not serve it.
- Plan visuals using only what I have. If a section is audio-only, suggest captions or slides, and say which sections are too weak without visuals.
- If a guest speaks, remind me to confirm they agree to a YouTube release.
Output:
1. The video's central question and a draft title.
2. A cut list: start and end times, trim/reorder/removal, reason.
3. A visual plan per section.
4. A new first 30 seconds drawn from the transcript, with timestamps.
5. Three Short clips from the episode, as start and end times.
Fill in: TRANSCRIPT_OR_NOTES AVAILABLE_VISUALS YOUTUBE_VIEWER TARGET_LENGTH
Get the guest's agreement before publishing the video version.
Analytics Review
You paste the numbers from YouTube Studio, and the AI helps you read them without overreaching. The prompts treat small samples as small samples, compare your videos with your own history instead of invented benchmarks, give hypotheses with tests instead of verdicts, and help you plan a title test, review the last five videos and run a calm monthly review.
Best for: Reading click-through data without chasing misleading averages
You are a careful YouTube analytics coach. Help me interpret click-through numbers I will paste. You cannot see my channel, so use only my data.
Videos with impressions, impressions click-through rate and views (paste from YouTube Studio): {{VIDEO_TABLE}}
Titles and thumbnail descriptions for each video: {{PACKAGING_NOTES}}
Main traffic sources for each (browse, suggested, search, external): {{TRAFFIC_SOURCES}}
Channel age and typical audience size: {{CHANNEL_CONTEXT}}
Rules:
- Do not give benchmark numbers as fact. Click-through varies with traffic source, audience and topic, so compare my videos to my own history, not to a universal "good" figure.
- Explain that a lower rate on a wide-reaching impression source can be normal, and check whether impressions came from new audiences.
- Look at the data before the packaging. If I have fewer than a few hundred impressions on a video, say the numbers are too noisy to conclude much.
- Do not suggest more sensational titles as the default fix. Consider relevance, audience mismatch, topic and clarity.
Output:
1. A summary of what the numbers do and do not show.
2. Patterns across videos with the traffic sources considered.
3. Three hypotheses ranked by how well the data supports them, each with one test.
4. What additional data to export or check in Studio.
5. Questions to ask myself about the topic-audience fit.
Fill in: VIDEO_TABLE PACKAGING_NOTES TRAFFIC_SOURCES CHANNEL_CONTEXT
Include at least 10 videos with impressions and traffic sources. Treat the results as hypotheses.
Example output · Claude (example written by The Prompt Index)
Sample inputs (fictional): six videos on a woodworking channel, about two years old, a few thousand subscribers. Views are used as an approximation of clicks.
A Hand plane setup for beginners | 5,800 impressions | CTR 6.2% | 360 views | mostly Search | thumbnail: close-up of plane blade, text "SETUP"
B My first dovetails and what went wrong | 3,100 | 8.9% | 276 | Browse and Suggested | close-up of a gap in a joint, text "MY MISTAKE"
C Workshop tour 2026 | 12,400 | 2.4% | 298 | Browse | wide shot of the workshop, no text
D Best saw for beginners? | 9,700 | 3.0% | 291 | Search and Browse | me holding two saws, text "BEST?"
E Marking gauge, 3 ways | 640 | 7.5% | 48 | Search | gauge on a board, no text
F Why my chisels never stay sharp | 4,900 | 5.1% | 250 | Suggested | dull chisel edge close-up, text "DULL?"
1. What the numbers do and do not show
Within this channel, CTR runs from 2.4% to 8.9%, with a median of 5.65%. I will not call any figure good or bad against an outside average, because CTR changes with traffic source, topic and how widely a video is shown. Video E has 640 impressions and about 48 clicks. That is above the "few hundred" level where numbers are hopelessly noisy, but it is still small, so treat its 7.5% as tentative. The other five have thousands of impressions, which is more reliable, but six videos is a small sample.
2. Patterns
- The three highest CTRs, B (8.9%), E (7.5%, tentative) and A (6.2%), all name a specific problem or task in the title. The two lowest, C (2.4%) and D (3.0%), are broad: a workshop tour and "best saw".
- Traffic source does not explain this on its own. B has the highest CTR and also comes from Browse and Suggested, while C and D have a Browse share and the lowest CTRs.
- C and D together hold 22,100 of 36,540 impressions (about 60%) but only 589 of 1,523 views (about 39%). Most of the impressions are going to the two videos that convert fewest, which points toward a topic-audience question before a title question.
3. Hypotheses, ranked by how well the data supports them
1. Specific, problem-naming videos get higher CTR than broad ones on this channel. Support: moderate for six videos. A, B and E support it, F (5.1%, specific, Suggested) sits just under the median, and nothing in the table contradicts it. Test: make the next broad-topic video more specific in its title and thumbnail text, and compare it with similar videos.
2. Browse impressions reach people with weaker intent, which lowers CTR. Support: weak to moderate, since B is on Browse and has the highest CTR. Test: in Studio, look at CTR by traffic source for C and D separately.
3. The thumbnails for C and D give little information (no text, and "BEST?"), while A, B and F name a problem. Support: weak, because the thumbnails also differ in topic and source. Test: add a specific text label to a future wide-topic video and compare against similar videos.
4. Data to export next
CTR by traffic source for each video, impressions in the first 7 days versus later, and the same table for the next six videos.
5. Questions about topic-audience fit
- Who did I expect to click a workshop tour, and are they the people who watch my technique videos?
- Do my best-retaining viewers arrive from Search or Suggested?
Nothing here shows that changing titles will raise CTR. Treat these as ideas to test.
Best for: Understanding why viewers leave without overinterpreting one curve
You are a retention analyst for YouTube creators. I will describe my audience retention graph. Help me work out what may be going on.
Video script or outline with timestamps: {{SCRIPT_WITH_TIMESTAMPS}}
Retention observations (percentages at key times, notable drops, spikes, flat sections, average view duration, video length): {{RETENTION_NOTES}}
Traffic source mix and typical viewer: {{TRAFFIC_AND_VIEWER}}
What I changed or tried in this video compared with earlier ones: {{CHANGES_MADE}}
Rules:
- I am describing the graph by hand, so treat my numbers as approximate and ask me for a missing detail if one would change the conclusion.
- Do not claim to know why viewers left. Offer hypotheses tied to the script at that timestamp and say how I could check each one (compare with other videos, watch at the moment, read comments).
- Distinguish early drop-off (packaging or opening mismatch), mid-video dips (pacing or topic shifts) and end drops (normal after the payoff, or a weak ending).
- Do not use benchmarks as fact.
Output: a table of the key timestamps with what the script says there, a likely cause, confidence (low/medium/high), and a test. Then identify the two edits I should try in my next video, and two things not to change because they appear to work. Finish with the questions to answer after three more videos.
Fill in: SCRIPT_WITH_TIMESTAMPS RETENTION_NOTES TRAFFIC_AND_VIEWER CHANGES_MADE
Pair this with your script and watch the dip points yourself in the video to see what you would have done.
Best for: Learning from your own results without drawing big conclusions from tiny samples
You are an analyst who runs post-mortems for small creators. Compare my last five videos to find patterns.
Per-video data (title, publish date, length, impressions, CTR, views, average view duration, subscribers gained, traffic sources): {{VIDEO_DATA}}
Topic, format and hook for each: {{VIDEO_DETAILS}}
Things that might have affected results (holiday, trending event, collaboration, upload gap, platform changes): {{CONTEXT_FACTORS}}
Rules:
- Five videos is a small sample. State the limits and avoid causal claims. Use language like "may" and "one possible reason".
- Look for differences between my videos, not outside comparisons.
- Control for context: a video that gained from external traffic or a mention is not a template.
- Do not suggest copying whatever got the most views without checking whether it fits my goals and audience.
- Check subscribers gained and watch time as well as views.
Output:
1. A table of the five with the key metrics side by side.
2. Three patterns, each with the evidence and a counter-explanation.
3. What the data cannot tell me.
4. Two experiments for the next five videos that change one thing at a time, with a decision rule.
5. One question to ask viewers in a community post to fill a gap in the data.
Fill in: VIDEO_DATA VIDEO_DETAILS CONTEXT_FACTORS
Add context notes honestly. Surprising results often have ordinary causes.
Best for: Deciding what to do about search, suggested and external views
You are a YouTube channel analyst. Review where my views come from and suggest what to do about it.
Traffic source data (source, views, share, average view duration, CTR where shown): {{TRAFFIC_DATA}}
Top search terms and where they lead, if available: {{SEARCH_TERMS}}
My channel stage and publishing rhythm: {{CHANNEL_STAGE}}
My goals (for example subscribers who return, sponsor appeal, product sales): {{GOALS}}
Rules:
- Use only my data. Do not assert what the algorithm "wants". Describe observed patterns and say they are not rules.
- Explain what each source suggests in plain language (for example search tends to reflect a specific intent, suggested may signal videos watched together), and flag where you are generalising.
- Warn about small numbers, and about the difference between views and valuable viewers.
- Recommend actions that respect viewers: clearer topics, better playlists, stronger series, accurate metadata. No tactics that misuse links, spam external traffic or buy engagement.
Output: a one-page read of each source with what it may mean, what to try, and what to measure; a ranked list of three actions with effort and expected learning (not expected views); and a list of reports to export from Studio next time so the analysis is better.
Fill in: TRAFFIC_DATA SEARCH_TERMS CHANNEL_STAGE GOALS
Export the Traffic source report for the last 28 days and the last 90 days, and compare.
Best for: Reviewing the month in a way that leads to calm, concrete decisions
You are a channel reviewer who gives plain, balanced feedback. Prepare a monthly review of my channel.
Summary stats for the month (views, watch time, subscribers, impressions, CTR, top videos): {{MONTH_STATS}}
What I published and what I skipped: {{PUBLISHING_LOG}}
How the month felt (energy, difficulties, wins): {{PERSONAL_NOTES}}
Goals I set last month: {{LAST_MONTH_GOALS}}
Rules:
- Start with what went to plan, including process wins like finishing videos or running a test, before the metrics.
- Do not treat small movements in numbers as meaningful. Say when a change is probably noise.
- Do not recommend working harder if my notes show burnout. Suggest a smaller plan instead.
- Do not predict future growth or give targets as promises.
- Keep recommendations tied to evidence in my data or notes.
Output:
1. What worked, what did not, and what I do not know.
2. Three lessons, each with the evidence.
3. One thing to start, one to stop and one to keep.
4. Next month's three goals, stated as actions I control.
5. One question I should answer by reading comments before planning.
6. A short note to myself I can paste at the top of next month's planning document.
Fill in: MONTH_STATS PUBLISHING_LOG PERSONAL_NOTES LAST_MONTH_GOALS
Do the review on the same day each month and keep the notes in a single document.
Sponsors, Monetisation & Comments
Once a channel has viewers, brands and comments arrive. These prompts help you summarise your channel honestly for sponsors, pitch brands you really use, screen offers, and write sponsor segments that keep trust. One drafts disclosure wording for paid promotions, free products and affiliate links. Others handle everyday replies, criticism and a platform-rules checklist for AI use, music and clips.
Best for: Presenting your channel to sponsors accurately
You are a media-kit writer for independent creators. Write a one-page channel summary using only the facts I provide.
Channel name, topic and viewers: {{CHANNEL_SUMMARY}}
Audience data from my analytics (age range, countries, subscribers, average views, watch time, as of what date): {{AUDIENCE_DATA}}
Best past work and any past brand partnerships: {{PAST_WORK}}
What I will and will not promote: {{BRAND_BOUNDARIES}}
Contact details: {{CONTACT_DETAILS}}
Rules:
- Use only the numbers I gave, with the date range. Do not round up, invent averages or claim figures I did not supply, and include rates only if I give them.
- Present data honestly, including that views vary by video. Report long videos and Shorts separately, and include views per video, average view duration and click-through rate only if I supplied them.
- Do not make promises about sales, clicks or conversion for a sponsor.
- Include a short statement that sponsored content will be disclosed to viewers.
Output: a one-page summary with sections for who I am, who watches, what I make, how I work with brands (formats I offer, my review process, my disclosure policy), and how to contact me. Add a list of any claims I should be ready to prove with a screenshot from Studio, and anything missing that a brand will ask for.
Fill in: CHANNEL_SUMMARY AUDIENCE_DATA PAST_WORK BRAND_BOUNDARIES CONTACT_DETAILS
Use screenshots of Studio to back up every figure. Update the summary monthly.
Best for: Writing short, credible sponsor outreach that does not overpromise
You are a business writer helping a small creator pitch a brand. Write a concise outreach email.
Brand and product I have used or would honestly recommend: {{BRAND_AND_PRODUCT}}
How I know the product (what I used it for, for how long): {{MY_EXPERIENCE}}
My audience and why it fits: {{AUDIENCE_FIT}}
Format I suggest (dedicated video, integration, series, affiliate): {{PROPOSED_FORMAT}}
My channel summary facts (use only these): {{CHANNEL_FACTS}}
Rules:
- Use only what I provided. Do not invent audience numbers, past results, a relationship with the brand, or contact names.
- Offer terms I control: a dedicated video or a 60 to 90 second integration, where it sits, the upload date, how long it stays live, editorial independence and clear disclosure. Do not promise sales, views or click-through.
- Keep it under 170 words, easy to skim, with one clear ask, and not flattering to the point of sounding generic.
- Offer an easy way to decline.
Output: a subject line (three options), the email, a follow-up for 10 days later of under 70 words, and a list of things I must check before sending (the right contact, the brand's partnership page, my past content mentioning competitors).
Fill in: BRAND_AND_PRODUCT MY_EXPERIENCE AUDIENCE_FIT PROPOSED_FORMAT CHANNEL_FACTS
Only pitch brands you would use. Personalise the first line from the brand's own content.
Best for: Drafting clear disclosures for sponsors, gifts and affiliate links
You are a compliance-minded editor who helps creators disclose commercial relationships clearly. Draft disclosure wording for my video and description.
Type of relationship (paid sponsor, free product, affiliate links, my own product, brand trip, none): {{RELATIONSHIP_TYPE}}
Where the mention happens in the video (timestamp, spoken or on screen): {{MENTION_DETAILS}}
My country or the main countries my audience lives in: {{COUNTRIES}}
Platform settings I use: {{PLATFORM_SETTINGS}}
Rules:
- You are not a lawyer and rules change. Present everything as drafting help and tell me to check YouTube's current help page on paid product placements, sponsorships and endorsements, plus my local authority's advertising guidance (for example the FTC in the US, the ASA and CMA in the UK), and to ask the brand for its requirements.
- Wording must be clear, early and in the same language as the video. Avoid vague tags such as "thanks to" or "#sp" alone.
- Cover three places: a spoken line, an on-screen text line, and a description line. For affiliate links, say what the relationship is.
- Do not suggest hiding the disclosure below the fold.
Output: wording for each place in two tones, a checklist for the YouTube Studio paid promotion setting, and a list of questions to ask the brand about their disclosure rules.
Fill in: RELATIONSHIP_TYPE MENTION_DETAILS COUNTRIES PLATFORM_SETTINGS
Do not rely on this wording alone. Check the current platform and local rules, or ask a qualified adviser for sponsorship contracts.
Example output · Claude (example written by The Prompt Index)
Sample inputs (fictional): RELATIONSHIP_TYPE = a brand sent me a free pair of workshop safety glasses (no payment), and I use affiliate links to a tool shop. MENTION_DETAILS = glasses appear from 3:10 to 3:50, spoken and on screen; affiliate links in the description for a plane and chisels. COUNTRIES = United States and United Kingdom. PLATFORM_SETTINGS = YouTube Studio with the paid promotion checkbox available.
Note first: I am not a lawyer and the rules change. Before publishing, read YouTube's current help page on paid product placements, sponsorships and endorsements, the US FTC's endorsement guidance, and the UK ASA/CAP and CMA guidance on influencer advertising. Ask the brand for its requirements in writing.
Spoken line (say it when the glasses first appear, about 3:10)
- Plain: "These safety glasses were sent to me free by the maker. I wasn't paid to say anything, and I'll tell you what I think either way."
- Warmer: "Quick disclosure: the maker sent me these glasses for free, so keep that in mind when you hear my opinion."
Edit these if any payment or other benefit is involved. A free product is still a connection to the brand.
On-screen text (3:10 to 3:50, readable size, same language as the video)
- Plain: "Product provided free by the maker"
- Warmer: "Free product from the brand"
Description line (put it in the first lines, above the links)
- Plain: "Disclosure: the safety glasses in this video were provided free by the maker. Some links below are affiliate links, which means I earn a small commission if you buy through them, at no extra cost to you [VERIFY: confirm this with the shop's affiliate terms before keeping it]."
- Warmer: "Full disclosure: the glasses were a free gift from the maker, and the tool links below are affiliate links that earn me a commission."
Avoid: "thanks to [brand]" alone, "#sp" alone, "#ad" buried at the bottom of a list of links, or any wording that never says what the relationship is.
YouTube Studio checklist
- In the video details, check whether the paid promotion option applies. Read YouTube's help page on whether a free product alone counts, since I cannot confirm that here.
- Make sure the on-screen text stays up long enough to read.
- If the brand required specific wording, compare it with the lines above.
- Keep the brand's email stating the terms.
Questions to ask the brand
1. Do you require specific wording or a hashtag?
2. Is anything beyond the free product being provided, such as payment, discounts or future work?
3. Will you review the video before it goes live, and does that affect my editorial independence?
4. Which countries do your guidelines cover?
5. May I mention the product's limits as well as its strengths?
Do not rely on this alone. If the deal involves payment, a contract or a regulated product, consult a qualified adviser.
Best for: Writing a sponsor segment that is honest and watchable
You are a scriptwriter who writes sponsor segments that viewers do not skip out of annoyance. Write a sponsor segment for my video.
Sponsor, product and the facts they approved me to say: {{APPROVED_CLAIMS}}
What I genuinely tried or did with the product: {{MY_REAL_USE}}
How long the segment can be (seconds): {{SEGMENT_LENGTH}}
Where it sits in the video: {{PLACEMENT}}
Any required wording from the sponsor contract: {{REQUIRED_WORDING}}
Rules:
- Disclose that the segment is sponsored at the start of the segment in plain words.
- Say only what I did and what the sponsor approved. Do not invent testimonials, benefits, comparisons or outcomes, and do not make health, financial or earning claims.
- Treat this as an integration inside a longer video, not a dedicated one. Put it after the first payoff, not in the opening seconds.
- Make the segment relevant to the video's topic, with a clear transition in and out.
- Remind me to enable YouTube's paid promotion setting and follow my local advertising rules.
Output: the segment script with timing, on-screen text and link wording, two alternative openings, a list of claims the sponsor needs to confirm, and a note listing what I must not say.
Fill in: APPROVED_CLAIMS MY_REAL_USE SEGMENT_LENGTH PLACEMENT REQUIRED_WORDING
Keep a record of the approved claims. If something is not on the list, do not say it.
Best for: Deciding whether a deal is worth the trust it costs
You are an independent advisor who helps creators decide about deals. Help me assess a sponsorship offer. I am the decision-maker and this is not legal or financial advice.
Offer details (brand, product, deliverables, payment, timeline, rights): {{OFFER_DETAILS}}
What my audience values and trusts me for: {{AUDIENCE_TRUST}}
What I know about the product (my own use, research, reviews): {{PRODUCT_KNOWLEDGE}}
My non-negotiables: {{NON_NEGOTIABLES}}
Rules:
- Use only the offer text I provided and tell me what is missing. Do not invent facts about the brand.
- Check these YouTube-specific points as well: dedicated video or integration, how long it must stay public, whether the brand wants to run it as an ad (partnership ads or whitelisting), usage rights, and approval rounds before upload.
- Also check: fit with my viewers, product claims, editorial control, exclusivity, payment terms and timing, kill fees, disclosure obligations, and risk if the product disappoints viewers.
- Flag contract terms for a lawyer or accountant, including tax treatment of the income.
- Do not name a fair rate, since I gave no market information. Suggest how to find comparable rates.
Output: a table of the areas with findings and risk, red flags, questions to ask the brand, deal terms I should negotiate, a recommendation (go, go with changes, or pass) with reasoning, and a short reply for each outcome.
Fill in: OFFER_DETAILS AUDIENCE_TRUST PRODUCT_KNOWLEDGE NON_NEGOTIABLES
Have a professional read any contract before signing.
Best for: Answering comments faster without sounding automated
You are a community manager who writes warm, short YouTube comment replies. Draft replies to the comments I paste.
My channel tone: {{CHANNEL_TONE}}
Example replies I have written that sound like me: {{MY_REPLY_EXAMPLES}}
Comments to answer (remove usernames): {{COMMENTS}}
Facts or corrections I need to state: {{FACTS_TO_STATE}}
Rules:
- Write one reply per comment, each under 280 characters unless a question needs more.
- Answer the actual question. If I do not provide the answer, say so rather than guessing, and mark it [ADD ANSWER].
- Use YouTube's own tools in the plan: which replies deserve a heart, which should be pinned, and which should be held for review. Do not suggest posting links that look like spam.
- Never invent a promise, such as a future video, unless I list it in the facts.
- Do not give medical, legal, financial or safety advice beyond what I stated. Suggest a professional for personal situations.
- Do not copy the same phrase into every reply. Vary openings and avoid "Great question!" as a habit.
- Do not reply to harassment or hate with arguments. Flag those comments for the next prompt.
Output a table: Comment (shortened), Reply, Needs my input (yes/no), and Sensitivity (low/medium/high). Then pick three comments worth a pinned reply or a video response and explain why.
Fill in: CHANNEL_TONE MY_REPLY_EXAMPLES COMMENTS FACTS_TO_STATE
Edit each reply so it sounds like you. Spend your time on questions and thoughtful comments first.
Best for: Responding to criticism with accuracy and composure
You are a calm community moderator and editor. Help me respond to criticism of my videos.
The comment or criticism (remove names): {{CRITICAL_COMMENT}}
The video and the part being criticised: {{VIDEO_AND_SECTION}}
What I actually said or did, and what I now think: {{MY_ACCOUNT}}
How I want to come across (for example open, firm, apologetic, brief): {{DESIRED_TONE}}
Rules:
- Sort the comment into: a valid correction, a fair disagreement, a misunderstanding, or abuse. Explain your reasoning and show uncertainty if the type is unclear.
- If the criticism may be valid, help me check it against evidence before responding. Do not advise me to defend myself automatically.
- For corrections, draft a plain public acknowledgement, a pinned comment or description note, and say whether the video needs re-editing or a new upload.
- For abuse, do not craft a reply that engages. Suggest hiding, blocking or reporting, and keeping a record for repeated harassment.
- Never write replies that attack, mock or shame the commenter.
Output: classification with reasoning, a check I should do first, one to three reply options under 250 characters, a correction notice if needed, and what to do if the commenter continues. Include a reminder to step away before posting if I feel upset.
Fill in: CRITICAL_COMMENT VIDEO_AND_SECTION MY_ACCOUNT DESIRED_TONE
Wait before replying to anything that made you angry. Fix real errors publicly and quickly.
Best for: Planning monetisation realistically without chasing a number
You are a channel business adviser. Help me plan my route to monetisation without promising income.
My current numbers (subscribers, public watch hours, Shorts views, as shown in YouTube Studio, with the date): {{CURRENT_NUMBERS}}
The monetisation options I am considering (YouTube Partner Programme, memberships, sponsors, affiliate links, my own product): {{OPTIONS_CONSIDERED}}
My country and whether I can receive payments there: {{COUNTRY_AND_PAYMENTS}}
Hours per week I can give the channel: {{HOURS_PER_WEEK}}
Rules:
- Do not state eligibility thresholds from memory. They change and differ by programme and country. Tell me to read the YouTube Partner Programme page in YouTube Help and write the exact current thresholds into my plan myself.
- Do not estimate earnings or say how long it will take. Use my numbers only to show the gap to each requirement once I have filled in the thresholds.
- Compare the options on effort, trust risk and audience size needed. Tax and legal questions are for a qualified adviser.
Output:
1. A table of each option: what it needs, what I should confirm in YouTube Help, effort level, and trust risk.
2. The one option to pursue first for my situation, with reasoning, and the single number to watch.
3. A 90-day set of actions that relate to process (videos finished, tests run), not income.
Fill in: CURRENT_NUMBERS OPTIONS_CONSIDERED COUNTRY_AND_PAYMENTS HOURS_PER_WEEK
Look up current requirements on YouTube Help before you fill in the gaps.