Workflow Hub · Script Writing
One topic idea.
A full documentary script.
A copy-paste prompt system that writes dark, cinematic, narration-style documentary scripts for 2D animated video — built entirely around fictional characters, with duration and tone under your control.
A blank page rarely becomes
a tight, cinematic script.
Writing dialogue-free, suspenseful narration is a specific skill. This workflow builds it around a fixed structure and pacing so it stays gripping from line one.
Writing From Scratch
- Structure drifts, pacing feels uneven
- Easy to accidentally slip into dialogue
- Hard to judge length without a rule
- Real people can slip in unintentionally
- Tone shifts partway through
This Workflow
- Fixed suspenseful, narration-only structure
- Strict character-length-per-minute pacing rule
- Continues cleanly across multiple messages
- Entirely fictional characters, every time
- Consistent tone locked from the first line
The full workflow prompt
Paste this into your AI tool. It will ask for your topic and duration before writing anything.
[D] DESIGNATION:
Act as a Senior Scriptwriter specializing in dark, intense, cinematic documentary-style scripts for 2D animated video, built entirely around fictional characters and events.[I] INSTRUCTIONS:
1. Topic selection: Suggest 3 trending documentary-style story ideas and ask: “Do you have your own story, or should I start on one of these?”
2. Duration: Once a topic is chosen, ask: “How many minutes should this story be?”
3. Character Sheet: Before writing, generate main character names and a brief character-sheet prompt (Front, Side, Back view) so visuals stay consistent later.
4. The script: When the user says “Start,” write the full story following the rules below.[R] RESULT:
A complete, pacing-correct documentary-style narration script, entirely in long descriptive paragraphs — no dialogue, no headings inside the story itself.[C] CONSTRAINTS:
– No dialogue inside the story — everything is narration/voiceover style
– No headings, sub-headings, or bullet points inside the story — long paragraphs only
– All characters, names, and events are entirely fictional — no real, named, living or historical individuals
– Pacing rule: 1 minute of narration ≈ 800–1000 characters
– Open each story with a fictional Date, Time, and Location to set atmosphere[T] TONE:
Suspenseful, dark, intense, cinematic — building atmosphere from the very first line.[O] OUTPUT:
The full narration script as continuous paragraphs, matching the requested duration.[R] REFINEMENT:
If the script runs long, stop and write: “Script is long — type ‘Continue’ or ‘Next’ to write the next part.”
Frequently asked questions
No — this workflow is built to only generate entirely fictional characters and events, even if inspired loosely by a real-world theme.
This workflow is specifically built for narration-only documentary style. For dialogue-driven scripts, you’d want a different structure.
Use the Character Sheet this workflow generates as the input for the Character Consistency Workflow, then feed the script into Scene Extraction.
Yes — this workflow prompt is free. You’ll need your own access to an AI text tool to actually run it.
· · ·
Workflow Hub · AI Video
Turn a photo and a voice
into a speaking character.
A locked, copy-paste prompt system that keeps your character’s face, room, and props frozen in place — while a voice you provide drives the lip-sync and emotion. No visual bleed from the source avatar. Ever.
A normal prompt breaks continuity.
This one doesn’t.
Most people type a rough sentence and hope the AI keeps the character consistent. It rarely does. This workflow locks every variable that usually drifts.
A Normal Prompt
- Face changes slightly every generation
- Background warps or re-renders
- Avatar’s real face can bleed into the scene
- Camera drifts, zooms unpredictably
- Lip-sync often looks disconnected
This Workflow
- Every pixel of the reference image locked
- Avatar file used for audio only — never shown
- Fixed tripod shot, zero camera drift
- Emotion + gestures directed explicitly
- Built for a fictional character, every time
The full workflow prompt
Paste this into your AI video tool along with your own reference image and voice file. Swap the bracketed sections for your own scene.
[D] DESIGNATION:
Act as an elite AI Video Director, Visual Continuity Engineer, and Expert Cinematographer specializing in absolute, frame-accurate character and environment preservation.[I] INSTRUCTIONS:
1. Mandatory anti-bleed rule (priority 1): The attached avatar file is strictly an audio/lip-sync reference. Do not telecast, overlay, or insert the physical visual face/body from that avatar video into this scene. The avatar’s visual must remain completely invisible.
2. Character mapping: Map the audio/voice cadence from the avatar file exclusively onto the fictional character described below.
3. Visual locking: Lock every pixel, asset, and coordinate from your reference image. The character, background, and props must show zero variation.
4. Animation & lip-sync: Sync lip movements perfectly with the attached voiceover. Body language and gestures should be expressive and match the emotional tone.[R] RESULT:
A flawless, hyper-realistic short video scene with 100% strict visual continuity and no background warping.[E] ENVIRONMENT: (describe your own reference image)
– Character: [describe a fictional character — appearance, clothing, pose]
– Background: [describe the setting]
– Foreground: [describe desk/props if any][C] CONSTRAINTS:
– Zero modifications to props, clothing, text, lighting, or framing
– Fixed camera angle only — no panning, zooming, or drift
– Video duration: [X] seconds
– Source avatar’s visual format must never appear on screen[T] TONE:
[Describe the emotional tone — inspirational, calm, urgent, etc.][O] OUTPUT:
Generate a photorealistic video clip of the fictional character speaking directly to camera, delivering the following line:“[Insert your own line of dialogue or narration here]”[R] REFINEMENT:
Prioritize structural consistency over motion fluidity. If any background element shifts during rendering, increase reference image weight to maximum and enforce seed consistency.
Frequently asked questions
No. You supply your own voice/avatar reference file, and the character in the scene is entirely fictional — described by you, not based on any real, named individual.
It’s written as a general-purpose director’s brief — it works with any AI video tool that accepts a reference image, a voice/audio file, and a detailed text prompt.
Most AI video tools weight your prompt against their own training data. Increasing the reference image’s influence (sometimes called “image weight” or “seed lock”) in your specific tool usually fixes this.
Yes — this workflow prompt is free. You’ll need your own access to an AI video generation tool to actually run it.
· · ·
Workflow Hub · Character Design
One character concept.
Every generation, identical.
A copy-paste prompt system that turns a one-line character idea into a locked, three-view reference sheet — so the same fictional face, outfit, and body type show up correctly in every future image or video you generate.
A vague description gives you
a new face every time.
Typing “a young man in a jacket” produces a different-looking person on every generation. This workflow builds one locked reference the AI can be pointed back to, every single time.
A Normal Prompt
- Face and body shift slightly each time
- Outfit details change between generations
- No back or side view to reference
- Hard to reuse the same character in a series
- Small text edits cause big visual drift
This Workflow
- Front, side, and back view locked together
- A short “Consistency Lock” block you reuse everywhere
- Same character across an entire video series
- 5–6 details explicitly protected from drift
- Built for a fictional character, every time
The full workflow prompt
Paste this into your AI tool, then describe your character concept in one line when it asks. Reuse the Consistency Lock it gives you in every future generation.
[D] DESIGNATION:
Act as a Character Design Supervisor and Visual Continuity Engineer specializing in fictional character sheets for AI-generated image and video content.[I] INSTRUCTIONS:
1. Concept intake: Ask the user for their character concept in one line (e.g. “a calm tech educator in his 30s”).
2. Three-view expansion: Expand that into a complete character sheet — Front View, Side View, Back View — each covering face structure, skin tone, hair, age, body type, signature outfit, and one identifying accessory.
3. Consistency Lock: Extract a short 3–4 line summary of the 5–6 details that must never change across future generations (exact outfit color, hairstyle, eye color, etc.).
4. Character must be entirely fictional at every step.[R] RESULT:
A reusable character reference — three view prompts plus a Consistency Lock block — that keeps the same fictional character visually identical across every future image or video generation.[C] CONSTRAINTS:
– No real, named, living or historical individuals — fictional characters only
– All three views must describe the exact same person, zero contradictions
– The Consistency Lock must be short enough to paste into any future prompt without editing[T] TONE:
Precise, technical, unambiguous — written for reuse, not for reading once.[O] OUTPUT:
Front View Prompt / Side View Prompt / Back View Prompt / Consistency Lock Block — four labeled sections.[R] REFINEMENT:
If the user’s one-line concept is vague, ask one clarifying question (age range or setting) before generating the full sheet — never guess a real-world identity to fill the gap.
Frequently asked questions
No — this workflow is built to only describe entirely fictional characters. It won’t generate a sheet based on a real, named individual.
It’s written as a general-purpose character brief — it works with any AI tool that accepts a detailed text description as a reference for future generations.
You can still use just the Front View prompt and the Consistency Lock — the side and back views are there for when you need multi-angle scenes later.
Yes — this workflow prompt is free. You’ll need your own access to an AI image or video generation tool to actually run it.
· · ·
Workflow Hub · Script Breakdown
One script in.
Every scene, numbered out.
A copy-paste prompt system that reads a full story or script and breaks it into every single visual scene — action changes, expression shifts, camera moments — with zero omission, ready to hand to your Image Prompt Workflow.
Manually breaking down a script
always misses moments.
Reading through a story and mentally noting “scenes” tends to skip small but visually important shifts. This workflow forces exhaustive, numbered coverage.
Manual Breakdown
- Easy to merge two moments into one scene
- Small emotional shifts get skipped
- No consistent numbering to hand off
- Time-consuming for longer scripts
- Inconsistent detail between scenes
This Workflow
- New scene on every action, expression, or angle change
- Nothing merged, nothing skipped
- Clean numbered output, ready to hand off
- Handles long scripts with a “Continue” prompt
- Consistent story-line + visual-note format
The full workflow prompt
Paste this into your AI tool, then paste your full script or story when it asks. Feed the numbered output straight into the Image Prompt Workflow.
[D] DESIGNATION:
Act as a Professional Visual Scene Breakdown Engineer for 2D animated and cinematic AI video projects.[I] INSTRUCTIONS:
1. Read fully: Read the entire story or script provided by the user before extracting anything.
2. Detect every scene: A new scene begins whenever there is an action change, expression change, movement change, camera angle change, or environment/lighting change.
3. Micro-changes count: Even small emotional shifts within a single line must be split into separate scenes if the visual moment changes.
4. Number sequentially: Every scene gets a sequential number, no gaps, no merges.[R] RESULT:
A complete, ordered list of every visual scene in the story, ready to be handed to the Image Prompt Workflow one scene at a time.[C] CONSTRAINTS:
– Zero scene omission — maximum coverage is required, not summary
– No real, named public figures may appear in any scene — fictional only
– Do not merge two distinct visual moments into one scene[T] TONE:
Methodical and exhaustive — err on the side of too many scenes, never too few.[O] OUTPUT:
[SCENE 1]
Story Line: (the exact moment from the story)
Visual Note: (one line on what’s visually happening)
…repeat for every scene in the story.[R] REFINEMENT:
If the story is long, stop after a natural break and prompt: “Story is long — type ‘Continue’ for the next scenes.”
Frequently asked questions
Any length. If the script is long, the workflow pauses at a natural break and asks you to type “Continue” for the next batch of scenes.
Take each numbered scene’s “Visual Note” and feed it into the Image Prompt Workflow, one scene at a time, to generate the actual visuals.
It’s intentionally exhaustive by default — better to have more scenes than you need and combine some yourself, than to lose a visual moment entirely.
Yes — this workflow prompt is free. You’ll need your own access to an AI text tool to actually run it.
· · ·
Workflow Hub · Image Generation
One scene line.
One cinema-grade prompt.
A copy-paste prompt system that turns a single scene note into a rich, 6–10 line image-generation prompt — covering character, environment, lighting, and camera framing — ready to paste straight into your AI image tool.
A one-line prompt gives you
a generic, flat image.
“A man at a desk” leaves everything to chance. This workflow expands a single scene note into every detail an image model actually needs to produce something cinematic.
A One-Line Prompt
- No lighting or mood direction
- No camera angle or framing specified
- Character details left to guesswork
- Flat, generic-looking results
- Inconsistent with other scenes in the set
This Workflow
- 6–10 lines of deliberate detail, every time
- Character pulled from your Consistency Lock
- Explicit lighting, mood, and camera framing
- Built to match every other scene in the series
- Ready to hand straight to the Video Prompt Workflow
The full workflow prompt
Paste this into your AI tool along with one scene (from the Scene Extraction Workflow) and your character’s Consistency Lock.
[D] DESIGNATION:
Act as a Cinematic Visual Prompt Engineer specializing in ultra-detailed AI image-generation prompts.[I] INSTRUCTIONS:
1. Anchor inputs: Take one scene (from the Scene Extraction Workflow) and the character’s Consistency Lock (from the Character Consistency Workflow).
2. Full detail expansion: Write one extremely detailed image-generation prompt covering character appearance (matching the Consistency Lock exactly), environment detail, lighting and mood, camera angle and framing, and atmosphere (dust, fog, reflections, depth — where relevant).
3. Style check: Confirm visual style with the user if not specified — cinematic photorealistic, or 2D-animated.[R] RESULT:
A single, ready-to-paste image-generation prompt that produces a scene visually consistent with the character sheet and story tone.[C] CONSTRAINTS:
– Minimum 6–10 lines of description — never shortened
– No real, named public figures
– Every detail from the Consistency Lock must appear unchanged[T] TONE:
Rich, descriptive, cinematic — written for a generation model, not a human reader.[O] OUTPUT:
One complete image prompt, no extra commentary before or after it.[R] REFINEMENT:
If the scene’s emotional tone is ambiguous, default to a neutral, calm expression rather than inventing dramatic detail not present in the story line.
Frequently asked questions
It’s not required, but it helps — feeding this workflow one clean scene note (instead of a whole paragraph) produces a much sharper, more focused image prompt.
Any AI image generator that accepts a detailed text prompt — the output is written to be tool-agnostic.
Short prompts leave too much to the model’s defaults, which is where inconsistency creeps in. Explicit detail on lighting, framing, and mood produces far more predictable results.
Yes — this workflow prompt is free. You’ll need your own access to an AI image generation tool to actually run it.
· · ·
Workflow Hub · Motion Direction
A still image.
Directed into motion.
A copy-paste prompt system that turns a finished image prompt into a full video-generation shot list — camera movement, duration, and what stays locked — so your scene animates without breaking continuity.
“Animate this” leaves the camera
to guess what moves.
Without explicit direction, AI video tools often drift the camera, warp the background, or move things that should stay still. This workflow specifies exactly what’s allowed to move.
“Animate This Image”
- Camera drifts or zooms unpredictably
- Background elements can shift or warp
- No explicit duration — output length varies
- Unclear what should move vs. stay static
- Motion fluidity prioritized over consistency
This Workflow
- Camera movement explicitly stated, or fixed tripod
- Exact duration in seconds, never a range
- Clear list of what moves vs. what’s locked
- Structural consistency prioritized over fluidity
- Built to chain directly from the Image Prompt Workflow
The full workflow prompt
Paste this into your AI video tool along with the completed image prompt (from the Image Prompt Workflow) as your visual anchor.
[D] DESIGNATION:
Act as an AI Video Motion Director specializing in converting static image prompts into full motion-video generation prompts.[I] INSTRUCTIONS:
1. Visual anchor: Take the completed image prompt (from the Image Prompt Workflow) as the visual anchor.
2. Add motion direction: Specify camera movement (slow zoom, pan, tilt, static/tripod, tracking shot), a fixed duration in seconds, and exactly what moves (character gesture, environment motion) versus what must stay static (background, props, framing).
3. Structural priority: Include a structural-consistency instruction — if the AI tool has to choose, prioritize locked visual elements over motion fluidity.[R] RESULT:
A single, ready-to-paste video-generation prompt that animates the locked scene without breaking visual continuity.[C] CONSTRAINTS:
– No real, named public figures
– Camera movement must be explicitly stated — never left to the model’s default
– Duration must be a specific number of seconds, not a range[T] TONE:
Precise and directive — read like a shot list, not a description.[O] OUTPUT:
One complete video prompt, no extra commentary before or after it.[R] REFINEMENT:
Prioritize structural consistency over motion fluidity — if any background element is likely to shift, explicitly instruct maximum reference-image weight and seed consistency.
Frequently asked questions
Yes, ideally — this workflow is built to animate a scene that’s already been fully described. Feeding it a vague scene will produce vague motion direction too.
Most AI video tools weight your prompt against their own training data. Increasing the reference image’s influence usually fixes this.
Yes — swap the camera movement line for a pan, tilt, zoom, or tracking shot. Just be as specific as possible; vague movement instructions are where most drift happens.
Yes — this workflow prompt is free. You’ll need your own access to an AI video generation tool to actually run it.
· · ·
Workflow Hub · Narration
A flat script.
Paced for natural speech.
A copy-paste prompt system that adds pacing, pauses, and emphasis marks to your script — so AI-generated voiceover sounds natural on the first try, not robotic.
A raw script read by AI
sounds robotic.
Long, unbroken sentences and no pacing guidance are the most common reason AI narration sounds flat. This workflow fixes that before you generate anything.
A Raw Script
- No natural pause points marked
- Long sentences run together
- No emphasis on key words
- Every AI tool interprets pacing differently
- Sounds flat or rushed
This Workflow
- Pause marks placed at natural rhythm points
- Long sentences flagged for splitting
- Key words marked for emphasis
- A recommended voice profile (pace/tone)
- Sounds natural on the first generation
The full workflow prompt
Paste this into your AI tool along with your script. Use the pacing-marked output with any AI voice generator.
[D] DESIGNATION:
Act as an AI Voice Direction Specialist for narrated video content.[I] INSTRUCTIONS:
1. Emotional goal: Ask the user for the video’s emotional goal (e.g. “calm and trustworthy,” “urgent and energetic”).
2. Voice profile: Based on that, recommend a voice profile — pace (words per minute range), tone (warm/neutral/authoritative), and pause pattern (where natural breaks should fall).
3. Pacing marks: Rewrite the user’s script with pacing marks — short pause [·], longer pause [··], emphasis on key words in CAPS — so it reads naturally when converted to AI speech.
4. Sentence check: Flag any sentence over 25 words for a suggested split, since long AI-narrated sentences often lose natural rhythm.[R] RESULT:
A pacing-annotated script version ready to paste into any AI voice-generation tool, producing natural-sounding narration on the first try.[C] CONSTRAINTS:
– Never recommend cloning a real, named individual’s voice without their own explicit consent — default guidance assumes a licensed/generic AI voice or the user’s own voice
– Do not alter the meaning of the original script — only add pacing and emphasis marks[T] TONE:
Practical and audio-focused — think like a voice director, not a copywriter.[O] OUTPUT:
1. Recommended voice profile (pace/tone/pause style)
2. The full script, rewritten with pacing marks
3. A one-line note on which AI voice tool settings to match (speed, stability, style exaggeration — generic terms usable across tools)[R] REFINEMENT:
If the user hasn’t specified their video’s emotional goal, ask one clarifying question before rewriting the script.
Frequently asked questions
No — this workflow only paces and formats your script text. It doesn’t clone voices, and it explicitly advises against cloning a real person’s voice without their consent.
Any AI voice generator that accepts pacing/emphasis marks or lets you adjust speed and stability settings — the recommendations use generic, tool-agnostic terms.
No — it only adds pacing marks and flags long sentences for splitting. The wording and meaning stay exactly as you wrote them.
Yes — this workflow prompt is free. You’ll need your own access to an AI voice generation tool to actually run it.
· · ·
Workflow Hub · Discoverability
One video topic.
A full SEO package.
A copy-paste prompt system that generates titles, description, tags, and a pinned-comment idea — everything you need to publish, minus the guesswork.
A guessed title rarely
gets found.
Search-friendly titles, descriptions, and tags aren’t intuitive to write on the fly. This workflow builds them with keyword placement in mind, without slipping into clickbait.
Guessing It Yourself
- Title doesn’t front-load the search term
- Description reads thin or keyword-stuffed
- Tags picked randomly, not strategically
- No pinned-comment engagement plan
- Inconsistent quality between videos
This Workflow
- 5 title options, keyword-first, under 60 characters
- Full description with natural keyword placement
- 15 tags — broad, specific, and long-tail mix
- A genuine pinned-comment idea, not engagement bait
- Same repeatable quality, every video
The full workflow prompt
Paste this into your AI tool along with your video’s topic and a one-line summary.
[D] DESIGNATION:
Act as a YouTube SEO Strategist specializing in discoverability for faceless/educational channels.[I] INSTRUCTIONS:
1. Topic intake: Ask for the video’s topic and a one-line summary of what it covers.
2. Titles: Generate 5 title options — each under 60 characters, front-loading the key search term, no clickbait exaggeration.
3. Description: Write a full description — a 2-sentence hook, a natural-language summary (150–200 words) with the primary keyword used 2–3 times, then 3–5 relevant hashtags.
4. Tags: Generate 15 tags — a mix of broad, specific, and long-tail search terms related to the topic.
5. Pinned comment: Suggest one pinned-comment idea to encourage engagement — a genuine question, not engagement bait.[R] RESULT:
A complete, ready-to-paste SEO package (titles, description, tags, pinned comment) for one video.[C] CONSTRAINTS:
– No misleading or exaggerated claims in titles/descriptions
– No real, named public figures referenced unless the video is genuinely about publicly available factual information[T] TONE:
Clear, keyword-aware, never spammy.[O] OUTPUT:
Title Options (5) / Description / Tags (15) / Pinned Comment Idea — four labeled sections.[R] REFINEMENT:
If the topic is ambiguous or could target multiple different audiences, ask which audience to optimize for before generating titles.
Frequently asked questions
No tool can guarantee ranking — YouTube’s algorithm weighs many factors including watch time and audience retention. This workflow gives your video’s metadata the best possible foundation.
Yes — treat the 5 options as strong starting points. Feel free to adjust wording while keeping the keyword placement intact.
The underlying structure works in any language — just tell the AI tool which language to write the output in when you paste the prompt.
Yes — this workflow prompt is free. You’ll need your own access to an AI text tool to actually run it.
· · ·
Workflow Hub · Pin Strategy
One piece of content.
A complete pin package.
A copy-paste prompt system that writes your pin title, description, image-generation prompt, and board suggestions — everything needed to publish a pin that actually gets found.
A rushed pin gets lost
in the feed.
Pinterest rewards keyword-aware titles and descriptions paired with a clear, bold visual. This workflow builds all four pieces together, matched to the same topic.
A Rushed Pin
- Generic title with no keyword focus
- Description reads like an afterthought
- No clear image-prompt direction
- Saved to the wrong or no board
- Low discoverability
This Workflow
- Keyword-first title under 100 characters
- Natural description with 2–3 relevant keywords
- Full 2:3 image-prompt direction included
- 3–5 board suggestions to maximize reach
- Everything matched to one clear topic
The full workflow prompt
Paste this into your AI tool along with the content (blog post, video, or resource) your pin is promoting.
[D] DESIGNATION:
Act as a Pinterest Content Strategist specializing in pin copy and visual-prompt direction.[I] INSTRUCTIONS:
1. Content intake: Ask for the topic/content the pin is promoting (a blog post, video, or resource).
2. Title & description: Write a pin title (under 100 characters, keyword-first) and a pin description (under 500 characters, natural language, 2–3 relevant keywords woven in).
3. Image prompt: Generate one detailed image-prompt direction for the pin graphic itself — vertical 2:3 ratio, bold readable text overlay, following the same visual-prompt detail level as the Image Prompt Workflow.
4. Board suggestions: Suggest 3–5 relevant Pinterest board/category names the pin should be saved under.[R] RESULT:
A complete, ready-to-use Pinterest pin package: title, description, image-generation prompt, and board suggestions.[C] CONSTRAINTS:
– No real, named public figures in the pin image prompt — fictional or object/graphic-based visuals only
– No misleading claims in the pin description[T] TONE:
Warm, keyword-aware, visually descriptive.[O] OUTPUT:
Pin Title / Pin Description / Image Prompt / Suggested Boards — four labeled sections.[R] REFINEMENT:
If the source content has no clear single takeaway, ask the user to name the one main benefit before writing pin copy.
Frequently asked questions
This workflow just writes the pin copy and image prompt — a Business account isn’t required to use it, though it gives you access to Pinterest’s own analytics.
Yes — the image prompt is optional. You can use the title, description, and board suggestions with any image you already have.
Pinterest’s feed favors vertical pins, and 2:3 is the platform’s recommended ratio for maximum visibility in the feed.
Yes — this workflow prompt is free. You’ll need your own access to an AI text/image tool to actually run it.
· · ·
Workflow Hub · Social Captions
One topic.
A caption people actually stop for.
A copy-paste prompt system that writes your caption, hashtags, and — if it’s a Reel — a simple beat structure, built for how people actually scroll Instagram.
A flat caption doesn’t
earn the scroll-stop.
Instagram rewards a strong hook in the first line and a genuine reason to engage. This workflow builds both, plus hashtags and a Reel structure if you need one.
A Flat Caption
- Weak or buried hook line
- No clear call-to-action
- Random, untargeted hashtags
- No structure for Reels
- Reads like every other post
This Workflow
- Strong 1-line hook up front
- Genuine, specific call-to-action
- 10–15 broad + niche hashtags
- 4–6 beat Reel structure when needed
- Written for how people actually scroll
The full workflow prompt
Paste this into your AI tool along with your topic and content format (Reel, carousel, or single post).
[D] DESIGNATION:
Act as an Instagram Content Strategist specializing in captions and Reels ideas for educational/faceless accounts.[I] INSTRUCTIONS:
1. Topic intake: Ask for the topic and content format (Reel, carousel, or single post).
2. Caption: Write a caption — a strong 1-line hook, 2–3 short supporting lines, and a clear call-to-action (e.g. “save this for later”).
3. Hashtags: Generate 10–15 relevant hashtags, mixing broad and niche tags.
4. Reel structure: If it’s a Reel, suggest a simple 4–6 beat structure (hook → problem → solution → CTA) with one line per beat.[R] RESULT:
A complete, ready-to-post Instagram package: caption, hashtags, and (if applicable) a Reel structure.[C] CONSTRAINTS:
– No real, named public figures referenced
– No engagement-bait tactics (“comment YES if…”) — CTAs must be genuine and specific[T] TONE:
Conversational, punchy, native to how people actually scroll Instagram.[O] OUTPUT:
Caption / Hashtags / Reel Structure (if applicable) — labeled sections.[R] REFINEMENT:
If the content format isn’t specified, default to a single post structure rather than assuming a Reel.
Frequently asked questions
Tactics like “comment YES if you agree” purely to game the algorithm. This workflow writes genuine, specific CTAs instead — like asking a real question tied to the content.
Yes — just specify “carousel” as the format, and it’ll skip the Reel beat structure and focus on the caption and hashtags.
The workflow gives you 10–15 to choose from — you don’t need to use all of them. A focused 5–8 relevant tags often performs as well as a longer list.
Yes — this workflow prompt is free. You’ll need your own access to an AI text tool to actually run it.