--- name: Video Blueprint category: content works_with: [claude, chatgpt, gemini] difficulty: plug-and-play version: 1.0 source: adapted (viral AI-video guide, rewritten; fake credentials and income framing removed) tested: no --- # Video Blueprint ## What it does Turns a video idea into a complete pre-production plan: hook, scene-by-scene script, character and location notes, camera directions, and ready-to-paste generation prompts — so the AI video tool executes a plan instead of improvising one. ## When to use it - Before opening any AI video tool (planning after generating is how you get 20 disconnected clips) - Your videos feel like clips, not a story - You're about to spend generation credits and want them spent on a plan ## The skill ``` Act as a director and screenwriter planning an AI-generated video. Rules: one main action per shot, every scene must earn its place in the story, and recurring characters keep identical descriptions in every prompt. No filler scenes. Video idea: [TOPIC / STORY] Length: [DURATION] Platform: [YOUTUBE / SHORTS / TIKTOK / INSTAGRAM] Audience: [WHO] Goal: [ENTERTAIN / EDUCATE / SELL] Tone and visual style: [E.G. documentary, noir, warm] Build the blueprint: 1. The concept in two sentences, and an opening hook for the first 5 seconds — the image or line that stops the scroll. 2. The story broken into scenes. For each scene: - Narration or dialogue (exact words) - Character(s): appearance, clothing, emotion, action - Location and atmosphere - Camera: shot type, angle, movement - Lighting - An image-generation prompt for the key frame - An animation instruction (the ONE action in this shot) - Estimated duration 3. The chronological shot list I can work through one by one in my video tool. 4. The consistency block: master descriptions for every recurring character and location, written once, to be pasted into every scene prompt verbatim. ``` ## Example output [TO FILL AFTER TESTING] ## Tweaks - Making shorts? Cap it at 6–8 scenes; more than that won't survive a 45-second cut - Run Character Consistency Guide next if you have more than one recurring character - Save the consistency block in a note — you'll paste it dozens of times --- --- name: Character Consistency Guide category: content works_with: [claude, chatgpt, gemini] difficulty: plug-and-play version: 1.0 source: adapted (viral AI-video guide, rewritten) tested: no --- # Character Consistency Guide ## What it does Builds a locked visual bible for your video's recurring characters and locations — one master description each — so your protagonist doesn't get a new face in scene four. Consistency is the difference between a story and a slideshow of strangers. ## When to use it - Any video with a character who appears in more than one scene - Your last video's "same" character changed hair, age, or outfit mid-story - Before generating anything, right after the blueprint ## The skill ``` Act as a film production designer responsible for visual continuity. Your output will be pasted verbatim into image and video generation prompts, so write descriptions as concrete visual facts, not vibes. My storyboard: [PASTE FROM VIDEO BLUEPRINT] For every recurring CHARACTER, write a master description: - Face: shape, distinctive features, skin, age - Hair: color, length, style - Build and posture - Clothing and accessories (exact items and colors — these never change unless the story requires it, and then note the change as its own locked variant) - Default expression and energy For every recurring LOCATION: - Architecture / environment specifics - Time of day and weather - Lighting character - 3 background elements that should appear every time - Style keywords Then produce: 1. One reference-image generation prompt per character and per location — for creating the anchor image I'll reuse. 2. The "paste block" for each: a compressed 2-3 line version of the master description to append to every scene prompt. 3. A continuity checklist: 5 things to eyeball in every generated shot before accepting it. ``` ## Example output [TO FILL AFTER TESTING] ## Tweaks - Generate the reference image FIRST, then use your tool's character/reference feature with it in every scene — the text block alone isn't enough - One story change (new outfit for act two)? Create it as a second locked variant, never freestyle it - Works for brand mascots and product shots, not just people --- --- name: Shot Prompt Builder category: content works_with: [claude, chatgpt, gemini] difficulty: plug-and-play version: 1.0 source: adapted (viral AI-video guide, rewritten) tested: no --- # Shot Prompt Builder ## What it does Converts one storyboard scene into one production-ready generation prompt: a single clear action, camera and lighting defined, character references attached, nothing improvised. The one-action rule is what keeps AI video from turning to soup. ## When to use it - For each scene, right before you generate it - A generation came back chaotic — rebuild the prompt instead of re-rolling - You're tempted to ask for "enters, sits, answers phone, argues, leaves" in one prompt (don't — that's five shots) ## The skill ``` Act as a cinematographer writing a generation prompt for an AI video tool. Hard rules: ONE main action per shot; nothing appears that isn't in the scene or references; movement must be physically plausible; if my scene actually contains multiple actions, split it and tell me. The scene: [PASTE FROM SHOT LIST] Character paste-block: [PASTE FROM CONSISTENCY GUIDE] Location paste-block: [PASTE FROM CONSISTENCY GUIDE] Write the shot: 1. THE ACTION — one sentence, one action. 2. Character: movement, body language, expression through the shot. 3. Camera: framing, angle, movement (or locked), and why it serves the emotion of this beat. 4. Lighting and atmosphere. 5. Background life: what moves subtly behind the action. 6. First frame → last frame: what the shot opens and closes on. 7. The final assembled generation prompt: references + action + camera + lighting in one paragraph, ready to paste. If this shot matters to the story (a hook, a reveal, an emotional beat), say "generate 2-3 variations of this one" — and what to compare between them. ``` ## Example output [TO FILL AFTER TESTING] ## Tweaks - Batch it: paste 3 scenes at once and get 3 shot prompts back — but still generate them one at a time - Rejected shots are data: note WHY (wrong motion? melted face?) and add that as a "never" line next time - The first-frame/last-frame pair matters most for tools that animate between keyframes --- --- name: Audio Plan category: content works_with: [claude, chatgpt, gemini] difficulty: plug-and-play version: 1.0 source: adapted (viral AI-video guide, rewritten) tested: no --- # Audio Plan ## What it does Builds the scene-by-scene sound design for your video — voice delivery, pauses, ambience, effects, music dynamics, and deliberate silence — as an editing timeline. Audio is where AI videos stop feeling like AI videos. ## When to use it - Visuals are generated, and you're about to record or generate voiceover - Your drafts feel flat even though the pictures are good (it's the audio) - Before the edit, so sound is designed rather than sprinkled ## The skill ``` Act as a sound designer and voice director. Principles: the voice carries the story, effects make the world believable, music carries emotion, and silence is a tool — recommend it where it beats sound. Write voice direction a non-actor can follow. The script with scenes: [PASTE] Voice tool I'll use: [ELEVENLABS / OTHER / RECORDING MYSELF] Overall mood: [DESCRIBE] For every scene: 1. The exact narration/dialogue, split per speaker — each character's lines separated so I can generate voices individually, never as one long file. 2. Voice direction: tone, pace, energy, and where the pauses go (mark them in the text with [pause]). 3. Ambience bed: the constant background of this scene. 4. Spot effects: max 3 per scene, tied to visible actions. 5. Music: mood, intensity, and where it rises, falls, or drops out. 6. Any moment where silence does the work. Then: the assembly timeline — scene by scene, the layer order and rough levels (voice / effects / ambience / music), plus the 3 places in the whole edit where audio should peak. ``` ## Example output [TO FILL AFTER TESTING] ## Tweaks - Generate voiceover scene by scene even when it's one narrator — pacing control beats convenience - Doing shorts? One ambience, one effect, one music decision. More gets muddy at phone volume - Foley beats library music for believability: footsteps, cloth, doors first, soundtrack last --- --- name: Video Retro category: content works_with: [claude, chatgpt, gemini] difficulty: plug-and-play version: 1.0 source: adapted (viral AI-video guide, rewritten; built out from a step the source only sketched) tested: no --- # Video Retro ## What it does Feeds a published video's real numbers back into the next one: where viewers left, what the drop-off points have in common, and the specific changes for the next production. The feedback loop the "post and hope" crowd skips. ## When to use it - 48–72 hours after publishing, when retention data exists - A video underperformed and you're about to guess why - Monthly, across everything published, to find patterns one video can't show ## The skill ``` Act as a video performance analyst. Rules: distinguish what the data shows from what you're hypothesizing — label hypotheses as hypotheses. No algorithm mythology ("the algorithm hates X") — stick to viewer behavior. If my data can't answer a question, say what data would. The video: [TITLE + one-line description + length] The numbers I have: [VIEWS / CTR / AVG VIEW DURATION / RETENTION GRAPH DESCRIPTION — e.g. "cliff at 0:45, slow slide from 3:00" / LIKES / COMMENTS / anything else] What happens in the video at the drop points: [DESCRIBE WHAT'S ON SCREEN AT THE TIMES RETENTION DIPS] Top 3 comments (verbatim): [PASTE OR "none"] My goal for this video was: [VIEWS / SUBS / LEADS / SALES] Give me: 1. The one-paragraph honest read: what the numbers say happened. 2. Drop-point diagnosis: for each dip, the most likely cause (hook debt, pacing, confusion, promise mismatch) — labeled confirmed-by-data or hypothesis. 3. What WORKED: the segments that held, and what they share. 4. Three specific changes for the next video, ordered by expected impact — each tied to a diagnosis above, not generic advice. 5. One thing to deliberately repeat unchanged, so I'm not reinventing what already works. 6. The single metric to watch on the next upload to test whether change #1 worked. ``` ## Example output [TO FILL AFTER TESTING] ## Tweaks - Keep a running log of the "three changes" across videos — after five videos the log is your personal playbook - Comments are retention data in disguise: "got boring in the middle" beats any graph - Run monthly across all videos with the same prompt for pattern-level findings --- --- name: The AI Video Pipeline category: content tools_required: [Any AI chat for planning, an AI video studio (e.g. Higgsfield Cinema Studio), an AI voice tool (e.g. ElevenLabs), a video editor (CapCut / Premiere / DaVinci Resolve)] cost: free tiers exist for every stage; paid tiers scale with volume time_to_set_up: one afternoon for your first planned video version: 1.0 source: adapted (viral "Claude + Higgsfield money machine" guide, rewritten; tools verified July 2026) tested: no --- # The AI Video Pipeline ## What it does Chains five skills into a film-production discipline for AI video: plan everything before generating anything, lock characters before scenes, build shot by shot, design the audio, then feed the numbers back into the next video. The AI chat is pre-production; the video studio is production; you are the director. ## What you need - An AI chat (any major one) for the planning skills - An AI video studio — Higgsfield's Cinema Studio is the one this guide was built around (verified real and current, July 2026); the pipeline works with any tool that supports character references. Version numbers churn fast, so pick features, not versions. - An AI voice tool and a normal video editor - The five skills: Video Blueprint, Character Consistency Guide, Shot Prompt Builder, Audio Plan, Video Retro ## Setup (one time) 1. **Decide what the videos are FOR before making any.** Channel, client work, or brand content — pick one. Not because a guide says so, but because the format, length, and platform choices in the Blueprint all depend on it. (Monetizing the skill itself is a different job — that's the Sell Something Pack, not this pipeline.) 2. **Run Video Blueprint** on your first idea → script, scenes, shot list, consistency block. 3. **Run Character Consistency Guide** → master descriptions, then generate the reference images in your video tool and save them. This is casting and set-building; skipping it is why AI videos fall apart. ## The workflow (per video) 4. **Shot by shot:** for each scene, run Shot Prompt Builder → generate the key frame with your references attached → animate → review against the continuity checklist → next scene. One action per shot, always. Generate 2–3 variations only for the shots that carry the story. 5. **Audio Plan** → generate voices scene by scene, layer ambience, effects, music per the timeline. 6. **Assemble** in your editor: visuals + the audio timeline. The plan means editing is assembly, not archaeology. 7. **Publish, wait 48–72 hours, run Video Retro** with the real numbers. Its "three changes" are the first input to the next video's Blueprint. ## Maintenance habits - Keep a prompt library: your best-performing Blueprint and shot prompts, reused and varied — never rebuilt from zero. - Keep the rejects log: every discarded generation gets a one-line reason; the log becomes your "never" list. - Re-check your video tool's feature set quarterly — this space changes faster than any guide, including this one. - Volume comes last: do not scale production until Retro shows people finish what you already make. ## What we stripped The source sold this as a "money machine" that "replaces an entire video production company" for "$0 — no excuses." Stripped: the free tiers are real but constrained, and a pipeline is not revenue. Every prompt in the source claimed "20+ years of experience" for the AI's persona — fake credentials, removed; the instructions work without the costume. The guide's tool claim was already stale on arrival ("Cinema Studio 2.5" — the current release is 3.0), which is why this rebuild names features instead of versions. The monetization sections (faceless-channel ad revenue, "your first $1") were dropped from this workflow: earning from video is covered honestly by the Sell Something Pack; ad-revenue projections for new channels are lottery math we don't republish.