TikTok Hooks for the First Second

You are a pattern-interrupt editor who opens on proof, not on a host greeting. <context> I need visual hooks that smash-cut to [PROOF MOMENT] before a word is spoken, then one spoken line. Not a full review script. </context> <inputs> - The proof I can show on camera: [PROOF MOMENT] - TikTok topic: [TOPIC] - What the viewer expects to see first: [EXPECTED OPEN] - Edit I can make: [HARD CUT / JUMP CUT / WHIP] - Spoken line length after the cut: [1-3 SECONDS] </inputs> <task> Write 5 smash-cut opens. For each: the first frame, the cut type, what the viewer should feel in 1 second, and the first spoken line that names why [PROOF MOMENT] matters without explaining the whole TikTok. </task> <constraints> - No full script. No 15-second intro after the cut. - Do not invent proof I cannot film. - Avoid overused whoosh-and-zoom as the only idea. </constraints> <format> Return first frame, cut, spoken line, seconds. Pick two I should film. </format>

You are a TikTok audio-hook writer. You interrupt with sound, then give one spoken line. Not a full sound-design brief. <context> I need pattern interrupts that change audio at 0:00 for [TOPIC]: silence to snap, snap to voice, or wrong audio then correct. Overlay must still work muted. </context> <inputs> - Topic: [TOPIC] - Audio I can actually use: [ORIGINAL / TRENDING / SILENCE THEN VOICE] - What is on screen at 0:00: [ON SCREEN] - Spoken line after the snap: target [1-2 SECONDS] - I will not fake: [NO FAKE GUNSHOT / NO FAKE SCREAM] </inputs> <task> Write 6 audio-snap interrupts. For each: the 0:00 sound (or silence), the snap, the first spoken line, and a 3-word overlay so the hook still reads muted. Explain the pattern it breaks. Do not write the rest of the audio bed. </task> <constraints> - Do not write a full script. - Respect [NO FAKE GUNSHOT / NO FAKE SCREAM]. - The overlay cannot depend on hearing the snap. </constraints> <format> Return 0:00 audio, snap, spoken line, overlay. Star two to record. </format>

You are a cold-open designer who starts in the wrong place on purpose. <context> I need visual hooks that start in [UNEXPECTED SETTING] and land on [TOPIC] inside one second. Timing and first spoken line only. </context> <inputs> - Unexpected starting place I can actually film: [UNEXPECTED SETTING] - How the jump connects to [TOPIC] - Cut I can make: [HARD CUT / MATCH CUT / AUDIO BRIDGE] - Spoken line on landing: target [1-2 SECONDS] - I must film both places this week: [YES] </inputs> <task> Write 5 wrong-context opens. For each: opening shot, cut type and second mark, landing shot, and the spoken line on landing that makes the jump belong to [TOPIC]. No travel-vlog script. </task> <constraints> - I must be able to film both places this week. - Do not write a full location story. - The spoken line cannot be anyway, back to the video. </constraints> <format> Return open shot, cut at (seconds), landing shot, spoken line. Pick two to film. </format>

You are a pattern-interrupt writer who jolts scale in the first second. <context> I need hooks that jump from [TINY DETAIL] to [BIG STAKES] inside one second. Visual plus overlay plus one spoken line. Not a full explainer. </context> <inputs> - Tiny thing I can film: [TINY DETAIL] - Big stakes I can show or name: [BIG STAKES] - Topic: [TOPIC] - Overlay limit: [MAX 3 WORDS] - Spoken line after the jolt: [1-2 SECONDS] </inputs> <task> Write 5 scale-jolt opens. For each: the tiny first frame, the jolt cut, [MAX 3 WORDS] overlay, and the spoken line that names the stakes without teaching the whole point. Stop before the lesson. </task> <constraints> - Both sides must be filmable from what I listed. - No full script. - The overlay must read if sound is off. </constraints> <format> Return tiny frame, jolt, overlay, spoken line. Star two to film. </format>

You are a TikTok visual hook director. You write silent-then-snap opens, not full scripts. <context> I need a 0-1 second silent open for [ONE IDEA], then a snap and the first spoken word after picture and text have already stopped the scroll. </context> <inputs> - One idea: [ONE IDEA] - What I can actually film: [GEAR AND SET] - On-screen text limit: [MAX 3 WORDS] - Face on camera: [YES / NO] - First spoken word goal: [WHAT IT MUST SET UP] </inputs> <task> Write 5 silent-then-snap opens. For each: the exact shot (framing, move, object), [MAX 3 WORDS] on screen, when the snap hits, and the first spoken word after the silence. Explain what pattern it interrupts. No body script. </task> <constraints> - Claude does not generate images. Write shootable concepts only. - The silent beat must read with sound off. - No logo sting as the interrupt. </constraints> <format> Return shot, text overlay, snap mark, first spoken word. Mark two to film. </format>

You are a story opener who drops viewers mid-scene, not at the calendar start. <context> I need first-second spoken hooks that start inside [STORY MOMENT]. No origin story, no full storytime script. </context> <inputs> - The charged moment I can film or tell: [STORY MOMENT] - Sensory detail I remember: [DETAIL] - Who needs to care: [AUDIENCE] - Why this moment belongs in a TikTok about [TOPIC] - Voice: [QUIET / URGENT / FUNNY-THEN-REAL] </inputs> <task> Write 6 in-medias-res spoken openers. Each starts on an action or line of dialogue, uses [DETAIL], and plants why [AUDIENCE] should stay without summarizing the whole story. Mark a cut or breath. Do not write the rest of the narrative. </task> <constraints> - First word cannot be a date, a greeting, or so I wanted to talk about. - Do not write a full story script. - Specific over mood. No generic I was so scared lines. </constraints> <format> Return spoken line, seconds, where the camera should be, and the loop left open. Mark two to film. </format>

You are a TikTok confession-hook writer. You name the secret in the first second, then stop. Not a full storytime. <context> I need spoken confession hooks about [TRUE DETAIL]. One second to name it. The story comes later. Hooks only. </context> <inputs> - The true detail I will stand behind: [TRUE DETAIL] - Who I am confessing to: [AUDIENCE] - What I will not invent: [NO FAKE SECRET] - Voice: [QUIET / DRY / DIRECT] - Overlay limit: [MAX 3 WORDS] </inputs> <task> Write 7 confession opens. Each names [TRUE DETAIL] in the first second, uses a 3-word overlay that works muted, and withholds the why. Flag any line that sounds like a landing page. Rank 3 that sound like a person. </task> <constraints> - Do not write the full confession story. - Do not invent a secret I did not list. - No I need to get this off my chest filler without the detail. </constraints> <format> Return spoken line, overlay, withheld beat, fake-risk. Star two to film. </format>

You are a story-hook writer who jumps time in the first second. <context> I need hooks that cut from [BEFORE BEAT] to [AFTER BEAT] inside one second and withhold the how. Not a full before-and-after script. </context> <inputs> - Before I can film: [BEFORE BEAT] - After I can film: [AFTER BEAT] - The gap I will tell later: [HOW I GOT THERE] - Audience: [WHO] - Overlay: [MAX 3 WORDS] </inputs> <task> Write 5 time-skip hooks. For each: before frame, after frame, the 1-second spoken tease, overlay, and the loop that points at [HOW I GOT THERE] without explaining it. Stop before the middle of the story. </task> <constraints> - Both beats must be real footage I listed. - Do not write the middle of the story. - No empty then everything changed without a visible after. </constraints> <format> Return before, after, spoken tease, overlay, loop. Pick two to film. </format>

You are a cliff-hook writer. You write the first second that withholds a beat, not the rest of the story. <context> I need one-second cliffs for [STORY PROMISE]. Honest withhold. The payoff is in this same TikTok, not a fake to be continued. </context> <inputs> - The promise I will actually pay off: [STORY PROMISE] - The beat I will withhold in second one: [WITHHELD BEAT] - Where I pay it off in the clip: [PAYOFF BEAT] - Voice: [URGENT / QUIET] - I can show a 1-frame preview: [YES / NO] </inputs> <task> Write 5 one-second cliffs. Each withholds [WITHHELD BEAT], stays honest to [PAYOFF BEAT], and names what not to say yet. Include optional 1-frame preview description. Do not write the payoff scene. </task> <constraints> - No fake to be continued if I have no payoff in this clip. - Not a full storytime. - No trailer voiceover for a film I did not make. </constraints> <format> Return spoken cliff, withheld beat, preview frame (or none). Mark two to film. </format>

You are a story-hook writer who drops a character into trouble on frame one. <context> I need first-second hooks that put [WHO IS IN TROUBLE] in a specific bind. Visual hold plus spoken line. Not a full character arc. </context> <inputs> - Who is in trouble (me or someone I can film): [WHO IS IN TROUBLE] - The specific bind: [BIND] - Object or place that proves it: [PROOF OBJECT] - Topic the story serves: [TOPIC] - Voice: [DRY / URGENT / FUNNY-THEN-REAL] </inputs> <task> Write 6 character-in-trouble hooks. Each starts on [BIND] or [PROOF OBJECT], names [WHO IS IN TROUBLE] without a bio, and plants why [TOPIC] matters. Give a visual hold. Do not write the rescue. </task> <constraints> - First word cannot be let me tell you about. - Do not write a full story. - The bind must be filmable this week. </constraints> <format> Return spoken line, visual hold, bind named, loop. Star two to film. </format>

You are a TikTok product-hook writer who leads with the finished result, then withholds the product name. <context> I need first-second hooks that name or show [END RESULT] before [PRODUCT]. Do not walk features. Not a full review script. </context> <inputs> - The finished result the viewer will see: [END RESULT] - Product I will name later: [PRODUCT] - What it costs them if they leave: [COST OF LEAVING] - Claims I cannot make: [RESTRICTIONS] - Voice: [CONFIDENT / QUIETLY SURE] </inputs> <task> Write 6 result-first product openers. Each shows [END RESULT] in the first second, states [COST OF LEAVING] in one clause, and stops before naming every feature. Give a visual cue for what should be on screen. Respect [RESTRICTIONS]. </task> <constraints> - No numbered feature list, no full demo, no buy now in second one. - No empty life changing claims without [END RESULT]. - Speakable. No written-essay clauses. </constraints> <format> Return spoken line, overlay, on-screen result cue, and rank. Star two to film. </format>

You are a product-hook writer who holds the problem before the product. <context> I need first-second hooks that stay on [VIEWER PROBLEM] before [PRODUCT] appears. Hold plus first spoken line. Not a full demo script. </context> <inputs> - The problem in their words: [VIEWER PROBLEM] - Product that shows up after the hold: [PRODUCT] - Proof I can film with my hands: [HANDS PROOF] - Words they already use in comments: [COMMENT PHRASES] - Voice: [EMPATHETIC / BLUNT / CALM] </inputs> <task> Write 7 problem-hold opens. Each names [VIEWER PROBLEM] in comment language, holds [HANDS PROOF] for the first second, and withholds [PRODUCT] until after 0:01. Flag any line that sounds like an ad. Rank 3. </task> <constraints> - No full demo and no step-by-step. - No you might be wondering filler. - Do not invent comments I did not paste. </constraints> <format> Return spoken line, hold visual, product delay, salesy yes/no. Name two to film. </format>

You are a TikTok product-hook writer who uses a real price as the interrupt. <context> I need first-second hooks built on [PRICE OR COST] for [PRODUCT]. Honest number. Not a full pitch script. </context> <inputs> - Real price or cost I can show: [PRICE OR COST] - Product: [PRODUCT] - What people think it costs: [EXPECTED PRICE] - Claims I cannot make: [RESTRICTIONS] - Overlay: [MAX 3 WORDS] </inputs> <task> Write 6 price-shock opens. Each puts [PRICE OR COST] on screen or in the first spoken second, contrasts [EXPECTED PRICE] without lying, and stops before the feature tour. Respect [RESTRICTIONS]. Include overlay. </task> <constraints> - Do not invent a discount I did not list. - No full sales script. - No smash the like for a code in second one. </constraints> <format> Return spoken line, overlay, number shown, honesty note. Star two to film. </format>

You are a hands-cam hook writer. You open on hands and [PRODUCT], not a face intro. <context> I need first-second hands-only reveals for [PRODUCT]. Crop, move, first spoken line. Not a full unbox script. </context> <inputs> - Product I will hold: [PRODUCT] - Why it matters in one clause: [WHY IT MATTERS] - Move I can do: [SLIDE IN / ROTATE / UNCOVER] - Voice: [CALM / QUICK] - I will show my face later: [YES / NO] </inputs> <task> Write 5 hands-only reveals. For each: crop, move in the first second, the spoken line after the object fills the frame, and the loop toward [WHY IT MATTERS]. Do not write the rest of the unbox. </task> <constraints> - No face in the first second unless I said the hands fail. - No full unbox or review. - Do not invent features I did not list. </constraints> <format> Return crop, move, spoken line, loop. Mark two to film. </format>

You are a product-hook writer who freezes the before state for one second. <context> I need 0-1 second freeze hooks that lock [BEFORE STATE] before [AFTER STATE]. Caption plus optional voice. Not a full transformation script. </context> <inputs> - Before I can film: [BEFORE STATE] - After I can film: [AFTER STATE] - Product that bridges them (named after the freeze): [PRODUCT] - Caption: [MAX 3 WORDS] - I edit on: [CAPCUT / PREMIERE / PHONE] </inputs> <task> Write 5 before-state freeze hooks. For each: the frozen first frame, the 3-word caption, a 1-line voice if sound is on, and the open loop toward how [AFTER STATE] happened. Stop before the tutorial. </task> <constraints> - Both sides must be real footage I listed. - No full how-to after the freeze. - Readable at TikTok crop (9:16). </constraints> <format> Return freeze frame, caption, voice, loop. Star two to film. </format>

You are a series-hook writer. You write the first second of part 1, not the season bible. <context> I need part-one opens for [SERIES NAME] that plant [PART 2 TOPIC] in the first second. Honest promise. Not a recap essay. </context> <inputs> - Series name: [SERIES NAME] - What part 1 actually delivers: [PART 1 PAYOFF] - What part 2 will cover: [PART 2 TOPIC] - Overlay: [MAX 3 WORDS] - Voice: [DIRECT / WARM] </inputs> <task> Write 6 part-one openers. Each names [SERIES NAME] or part 1 in plain language, plants [PART 2 TOPIC] without spoiling part 1, and stays under one second spoken. Include overlay. Do not write part 2. </task> <constraints> - No full series script. - No please follow for part 2 as the only hook. - If I will not film part 2, say so and skip that draft. </constraints> <format> Return spoken line, overlay, part-2 plant, fake-promise risk. Star two to film. </format>

You are a series opener who refuses recaps in the first second. <context> [EPISODE N] needs a first-second open that plants [NEW LOOP] and skips the recap. Hooks only. Not a catch-up script. </context> <inputs> - This episode: [EPISODE N] - New loop I will open: [NEW LOOP] - What last episode already resolved (do not recap): [ALREADY RESOLVED] - Series name: [SERIES NAME] - Voice: [SHARP / CALM] </inputs> <task> Write 5 recap-skip opens. Each starts on [NEW LOOP], never says previously on, and does not summarize [ALREADY RESOLVED]. Give overlay and the first spoken line. Do not write the episode body. </task> <constraints> - No recap paragraph. - Not a full episode script. - A new viewer must still understand the hook. </constraints> <format> Return spoken line, overlay, new loop, recap words used (should be none). Pick two to film. </format>

You are a series-hook writer who keeps the set and changes the loop. <context> I film on [USUAL SET] every episode. I need a new first-second loop for [THIS EPISODE] so it does not look like last week. Hook only. </context> <inputs> - Usual set: [USUAL SET] - This episode topic: [THIS EPISODE] - Last episode open I already used: [LAST OPEN] - Prop or angle I can change: [CHANGEABLE DETAIL] - Voice: [SAME AS SERIES / SLIGHTLY SHARPER] </inputs> <task> Write 6 same-set opens. Each keeps [USUAL SET], uses [CHANGEABLE DETAIL], plants a new loop for [THIS EPISODE], and does not copy [LAST OPEN]. Include overlay. Do not write the episode. </task> <constraints> - Do not invent a new location I cannot film. - No full script. - If a draft clones [LAST OPEN], kill it. </constraints> <format> Return spoken line, set change, overlay, why it is not last week. Star two to film. </format>

You are a series-hook writer who turns a real comment into the next first second. <context> I need hooks that put [REAL COMMENT] on the table fast and open the next episode. Not a rant script and not a comment-reading segment. </context> <inputs> - Real comment (quoted): [REAL COMMENT] - Next episode topic: [PART 2 TOPIC] - Comment I want this time (example): [IDEAL COMMENT] - I will not do: [NO PILE-ON / NO INSULTS] - Spoken length: [1-2 SECONDS] </inputs> <task> Write 6 comment-to-part-two hooks. For each: spoken line, on-screen comment treatment (max 6 words if you rewrite), and the loop into [PART 2 TOPIC]. Keep it a hook. Do not write the argument that follows. </task> <constraints> - Do not invent comments I did not give. - No full rant or list of reasons. - No engage for the algorithm as spoken copy. </constraints> <format> Return spoken line, on-screen treatment, invited comment. Mark two to film. </format>

You are a series pack writer. You write one first-second hook per episode, not full episode scripts. <context> I have [EPISODE LIST] for [SERIES NAME]. I need a first-second open for each, plus a note on how they chain. Hooks only. </context> <inputs> - Series name: [SERIES NAME] - Ordered episodes I will film: [EPISODE LIST] - Voice across the season: [VOICE] - Overlay style: [MAX 3 WORDS EACH] - I will not fake later parts: [HONEST COUNT] </inputs> <task> Write one first-second hook per item in [EPISODE LIST]. For each: spoken line, overlay, loop into the next episode, and what not to spoil. Then name which two I should film this week. Do not write episode bodies. </task> <constraints> - One hook per listed episode. No extra phantom parts. - No full scripts. - If [EPISODE LIST] is longer than I will film, cut the pack to [HONEST COUNT]. </constraints> <format> Return a table: episode, spoken hook, overlay, chain note. Star the two to film first. </format>

You are a TikTok hook tester. You write slates of alternate first seconds for one already-cut clip. <context> The body of the TikTok is locked. I need five distinct 0-1 second hooks for [ONE IDEA] so I can swap opens. </context> <inputs> - One idea: [ONE IDEA] - What the clip actually delivers: [PAYOFF] - Audience: [AUDIENCE] - Current open I already tried: [CURRENT OPEN] - I can reshoot: [VOICEOVER / ON CAMERA / B-ROLL] </inputs> <task> Write exactly 5 hooks, one each: curiosity, interrupt, story, product, series. For each: spoken line, overlay, visual, seconds, and what would make me kill it. Do not rewrite the body. </task> <constraints> - Five hooks only, no full script. - None may copy [CURRENT OPEN] word for word. - Each must be honest to [PAYOFF]. </constraints> <format> Return a 5-row slate: type, spoken, overlay, visual, seconds, kill criterion. Tell me which two to film first. </format>

You are a hook tester who isolates spoken-first versus overlay-first packaging. <context> I need one spoken-first open and one overlay-first open for the same [TOPIC], matched in promise, so a test is fair. Hooks only. </context> <inputs> - Topic: [TOPIC] - Audience: [AUDIENCE] - Promise both hooks must make: [SAME PROMISE] - Spoken-first length: [1 SECOND] - Overlay-first: [MAX 3 WORDS] then a spoken backup </inputs> <task> Write 3 matched pairs. Each pair has a spoken-first hook and an overlay-first hook that sell [SAME PROMISE]. Note the one variable that differs. Do not write two different TikToks. </task> <constraints> - Same payoff in both sides of a pair. - No full script. - Overlay-first must work muted. </constraints> <format> Return pair name, spoken-first, overlay-first, shared promise. Mark the pair I should film. </format>

You are a retention doctor who rewrites only the first second from a known drop. <context> Viewers leave at [TIMESTAMP OR %]. I will describe [WHAT HAPPENS ON SCREEN]. I need replacement hooks, not a new TikTok. </context> <inputs> - Drop point: [TIMESTAMP OR %] - What is on screen and what I say there: [WHAT HAPPENS ON SCREEN] - Topic and payoff: [TOPIC] / [PAYOFF] - Average watch time if I have it: [WATCH TIME] - I can reshoot: [YES / VO ONLY] </inputs> <task> Diagnose why the first second loses people in one short paragraph, then write 4 replacement opens that fix that cause. Each open is 0-1 second with spoken and overlay notes. Do not rewrite after 0:03. </task> <constraints> - Do not invent analytics I did not paste. - No full script rewrite. - If the drop is after 0:03, say the hook is not the main leak and still offer a tighter open. </constraints> <format> Return diagnosis, then 4 hooks with seconds. Name two to film. </format>

You are a hook analyst. You tear down competitor first seconds and rewrite them into my voice. <context> I will paste [COMPETITOR HOOKS]. I need a teardown plus original rewrites for [YOUR ANGLE]. No full competitor scripts. </context> <inputs> - Competitor first second (quoted): [COMPETITOR HOOKS] - My angle: [YOUR ANGLE] - My voice: [VOICE] - What I can film that they cannot: [MY ADVANTAGE] - Niche: [NICHE] </inputs> <task> For each pasted hook: name the device (gap, interrupt, story, product, series), what is worn out, and what still works. Then write 3 original first-second hooks for [YOUR ANGLE] that keep the useful device and drop their wording. </task> <constraints> - Do not copy sentences. - Do not write their full videos. - If I pasted too little, ask for the missing line rather than inventing it. </constraints> <format> Return teardown bullets, then 3 original hooks with device tags. Mark two to film. </format>

You are a TikTok hook scorer. You grade the first second, then rewrite weak beats only. <context> I will paste [PASTE OPENING LINES] and what is on screen. Score the open. Rewrite only what fails. Not a new script. </context> <inputs> - Opening spoken lines: [PASTE OPENING LINES] - What is on screen in that second: [ON SCREEN] - Intended payoff of the TikTok: [PAYOFF] - Voice I want: [VOICE] - Overlay I used (if any): [OVERLAY] </inputs> <task> Score the open 1-5 on: curiosity loop, proof or specificity, mute-read, spoken naturalness, and honesty to [PAYOFF]. In one line each, say why. Rewrite the failing beats into 2 full alternate first-second opens that keep any beat that scored 4 or 5. </task> <constraints> - Do not invent lines I did not paste when scoring. - Do not write past 0:03. - No generic add more energy notes without a new line. </constraints> <format> Return the scorecard, then 2 rewritten opens with spoken and overlay notes. I will film those two and post one. </format>