[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7AMF0FWJ9319X7GK99G1VT5":3,"skill-same:01M7AMF0FWJ9319X7GK99G1VT5":48},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":27,"repositoryHtmlUrl":28,"repositoryHomepageUrl":29,"ownerHtmlUrl":30,"ownerBlogUrl":31,"ownerDescription":32,"updatedAt":33,"install":34,"resources":47},"01M7AMF0FWJ9319X7GK99G1VT5","pipeline-ai-film-production",null,"Use for Phases 10–13 of the Content Engineering Pipeline: designing and producing the AI-generated hub film from a locked screenplay (Phase 10-11) and then designing and producing the five-spoke content distribution campaign (Phase 12-13). Triggers on AI video generation, hub-and-spoke content model, transmedia distribution, spoke architecture, TikTok\u002FLinkedIn\u002Femail content strategy, or AI film production with Sora\u002FElevenLabs\u002FKling.","the-vibey-project\u002Fvibey","the-vibey-project","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F329298758?v=4","01M7AJHGZ9722CTJ4JNQZVRH0Q","the vibey project","01M7AJHGZVMG2K7TGXQ6XFHVG3",5,"MIT","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fcontent-engineering-pipeline\u002Fskills\u002Fpipeline-ai-film-production\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: pipeline-ai-film-production\ndescription: \"Use for Phases 10–13 of the Content Engineering Pipeline: designing and producing the AI-generated hub film from a locked screenplay (Phase 10-11) and then designing and producing the five-spoke content distribution campaign (Phase 12-13). Triggers on AI video generation, hub-and-spoke content model, transmedia distribution, spoke architecture, TikTok\u002FLinkedIn\u002Femail content strategy, or AI film production with Sora\u002FElevenLabs\u002FKling.\"\n---\n\n# Phases 10–13: AI Film Production and Spoke Campaign\n# Content Engineering Pipeline\n\n---\n\n## OVERVIEW: WHAT THESE FOUR PHASES ACCOMPLISH\n\nPhase 10 (Hub Architecture) → Phase 11 (Hub Draft) → Phase 12 (Spoke Architecture)\n→ Phase 13 (Spoke Drafts)\n\nThese four phases take the locked Phase 9 screenplay and produce: a complete\nAI-generated hub film (Phases 10–11), then a five-platform spoke content campaign\ndesigned to drive discovery of the hub film (Phases 12–13).\n\nThe hub film is the third conversion in the pipeline. The first conversion\n(Phases 6–7) moved from non-fiction research book to novel. The second\n(Phases 8–9) moved from novel to screenplay. This conversion moves from\nscreenplay to produced film — and it introduces the AI tool stack.\n\n---\n\n## PHASE 10: HUB ARCHITECTURE\n\n### What the Hub Film Is and What It Must Do\n\nThe hub film serves two simultaneous functions that its production must honor\nsimultaneously. As a standalone film, it delivers the controlling idea the\npipeline has been building since Phase 1. As a content architecture anchor,\nit is the pillar asset from which the spoke system of Phases 12–13 will extract,\nreformat, and distribute across five platforms.\n\nGary Vaynerchuk's reverse pyramid model requires a pillar content piece dense\nenough in extractable material to yield 15 or more derivative pieces. For a hub\nfilm, this means scenes must be designed as modular sequences: self-contained\nenough to produce emotional impact without full narrative context, yet specific\nenough to drive curiosity toward the complete hub experience. Every significant\nscene must be designed with both its in-film function and its out-of-film\nextractability in mind simultaneously.\n\nA film that works only as narrative and not as extractable content is not a hub.\nA collection of extractable moments that does not cohere as a film is not a hub.\n\nJonah Berger's research establishes that stories are the optimal sharing vehicle —\n\"Trojan Horses\" carrying information under the guise of narrative — and that\nhigh-arousal emotions, particularly awe, are the single strongest driver of virality.\nFor awe moments to be extractable as spokes, they must be designed as modular\nsequences with both in-film and out-of-film function.\n\n### The AI Tool Stack: Strengths and Failure Modes\n\nThe AI tool stack available for hub production includes tools with specific\nperformance profiles that make each optimal for specific scene types. These are\nnot interchangeable.\n\n> ⚠️ **Volatile facts (as of September 2026 — verify before relying on it).** Every\n> tool name, model version, capability, duration limit, resolution, partnership and\n> price in this section is a dated claim about a fast-moving market: Sora 2 (25-second\n> clips, up to 4K, native 9:16, synchronized audio, the Disney partnership), ElevenLabs\n> (inline emotional tags, Sound Effects v2), Nano Banana 2 (Google Gemini 3.1 Flash\n> Image: up to 4K, text rendering, character consistency, $0.039 per image), Kling AI\n> 2.0 (120-second single generations), MuseTalk 1.5, Wav2Lip-HQ and Google Veo 3.\n> Check each vendor's current documentation and pricing before planning a production\n> around it.\n\n**Sora 2** generates clips up to 25 seconds at up to 4K resolution in both 16:9\nand native 9:16 formats, with synchronized dialogue, sound effects, and music.\nPrimary tool for scenes where visual spectacle, environmental atmosphere, or\naction is the dominant element. Its Disney partnership production-grade positioning\nmakes it the highest-quality option for scenes where the visual world carries the\nargument. Failure mode: attempting multiple simultaneous choreographic events in\na single generation produces unreliable results. One camera move plus one subject\naction per generation maximum.\n\nThe image-to-video workflow is the preferred method for every extraction point\nscene. Generate the exact frame from Nano Banana 2 first, then animate that\nreference frame using Sora 2's image-to-video mode. Text-to-video produces\ngeneral results from descriptions; image-to-video produces specific results from\na designed source frame.\n\n**ElevenLabs** handles every character voice in the film regardless of which\nvisual tool is used for the footage. Audio consistency standard: using ElevenLabs\nfor all voice generation means all character voices share the same production\nquality register. Supports inline tags for emotional control — [softly], [whispers],\n[with restrained contempt], [voice breaking] — derived from the Phase 8 Coppola\nfive-category analysis. Failure mode: generating dialogue with neutral delivery and\nattempting to correct emotional register in post-production. Specify the delivery\nin the prompt.\n\nElevenLabs also handles ambient room tone generation (Sound Effects v2): the\nspecific acoustic character of each physical space layered under character dialogue.\nThis is the AI equivalent of Ben Burtt's \"worldizing\" technique. Prompt for room\ntone must name the specific space and its acoustic qualities concretely: not\n\"interior ambience\" but \"small kitchen, low ceiling, tile floor, fluorescent hum,\nslight refrigerator vibration.\"\n\n**Nano Banana 2** (Google Gemini 3.1 Flash Image) generates still assets at up to\n4K resolution with accurate text rendering and character consistency, at $0.039 per\nimage. Primary tool for title cards, reference images, atmospheric stills, and any\nin-frame visible text. Its text accuracy far exceeds the video generation tools.\n\n**Kling AI 2.0** supports up to 120 seconds of single-generation duration with\nstrong facial performance quality. Primary tool for scenes where sustained character\npresence and emotional expressiveness are the dominant requirements. Its longer\ngeneration window allows full scene coverage without cut-assembly work. For dialogue\nscenes, use the video-first then lip-sync workflow: generate character performance\nvisually using Kling AI, then apply MuseTalk 1.5 or Wav2Lip-HQ to synchronize the\nseparately-generated ElevenLabs audio to the generated video.\n\n**Google Veo 3** provides the highest quality audio-visual synchronization across the\ntool stack. Optimal choice for extraction point scenes where the emotional peak\ndepends on the specific quality of a character's delivery synchronized precisely with\ntheir visible emotional state — typically the All Is Lost scene, the Break into Three\nmoment, and the Closing Image.\n\n### Pre-Production Design Work\n\nTwo Phase 8 documents feed directly into Phase 10 work:\n\n`08_spine_and_anchors.md` and `08e_visual_grammar.md` contain the visual grammar\ndesigned for each primary location and the compression character designs that defined\neach character's visual signature. These become production design specifications: visual\ngrammar notes are translated from cinematic intention into prompt-engineering language,\nand character visual signatures become the reference image descriptions that ensure\nconsistency across multiple AI generations.\n\nFor each primary location, the visual identity specification must name at least one\nspecified imperfection. Generation tools default to pristine, hyper-detailed, slightly\nsynthetic aesthetics — every surface looks new, every texture looks described. George\nLucas and art director Roger Christian built Star Wars' visual authority by making\neverything worn, used, and imperfect — the \"used future\" aesthetic — which made a\nscience fiction world feel inhabited rather than designed. The prompt vocabulary for\nany scene that should feel real must include imperfection details explicitly, because\nthe generation model will not add them unless specified.\n\nThe Prompt Lexicon (`10f_prompt_lexicon.md`) is the most important production tool\nfor Phase 11 consistency. It contains Constant Character Prompts (one verbatim\nparagraph per major character) and Constant Style Prompts (one block per primary\nlocation). The discipline it enforces is absolute: no generation prompt is written\nfrom memory. The character description that goes into every prompt is the same text\nthat went into every prior prompt, copied from the same source. Any variation in\ncharacter description language produces variation in generated appearance.\n\n### Phase 10 Production Design Documents\n\n**10_extraction_point_map.md** — The hub film's content architecture blueprint.\nLists every extraction point organized by tier (Tier 1: inciting incident, All Is\nLost, most quotable dialogue, character's defining decision, most visually iconic\nsequence; Tier 2: midpoint reversal, key conflict scenes, thematic revelations;\nTier 3: production design details, character introductions, tonal moments). For\neach point: page reference, self-contextualizing description, spoke assignments,\n9:16 generation flag, curiosity cascade logic (what fragment does this reveal and\nwhat 30–40% does it withhold), and Interest Market Cap framing (high-reach framing\nof the curiosity gap rather than clinical framing).\n\n**10b_visual_identity_spec.md** — Full prompt vocabulary for every primary location:\nlighting temperature and direction, color palette with specific reference descriptors\n(not just \"warm\" but \"amber tungsten warmth, shadows cool blue-grey\"), texture and\ngrain character, camera-to-subject spatial relationship, recurring motif detail stated\nin visual terms a generation model will render consistently, and at least one specified\nimperfection per location.\n\n**10c_character_reference_lib\u002F** — One subfolder per major character containing the\nwritten reference description in prompt-ready language, the voice profile specification,\nElevenLabs parameters, and three to five Nano Banana 2-generated reference images.\nReference images are the Phase 10 most important technical output — they are generated\nduring Phase 10, not Phase 11, because evaluating their adequacy as consistency anchors\nis an architecture decision, not a production decision.\n\n**10d_tool_assignment_map.md** — Every scene with tool assignment, justification based\non the scene's specific requirements, and generation estimate. Also includes the\ngeneration prompt template for each scene (not the final prompt, but the template\nestablishing what the prompt must contain: character reference, location and visual\nidentity specification, action description, emotional register, technical requirements).\n\n**10e_technical_specs.md** — All production parameters: 4K resolution (3840×2160),\n24fps, ProRes or H.265, separate audio stems for dialogue, sound effects, and music.\nAspect ratio decision (16:9 primary, 9:16 for flagged extraction points). Reformatting\npipeline, caption generation standard (97% minimum), UTM parameter convention.\n\n**10f_prompt_lexicon.md** — Copy-paste production document. Constant Character Prompts\nand Constant Style Prompts formatted for copy-paste use during generation sessions.\nKept open during all Phase 11 generation sessions.\n\n### Phase 10 Quality Gate\n\nPhase 10 is complete when: the extraction point map lists a minimum of six primary\nextraction points with curiosity cascade logic and Interest Market Cap framing; the\nvisual identity specification covers every primary location including specified\nimperfections; the character reference library has a complete entry with generated\nreference images for every major character; the tool assignment map covers every scene;\nthe technical specifications sheet is complete; the Prompt Lexicon is assembled and\nverified against reference images; and the Hub Fidelity standard has been applied to\nevery extraction point.\n\n---\n\n## PHASE 11: HUB DRAFT\n\n### Production Sequence\n\nProduce in this order:\n\n1. Phase 10 gate confirmation — all five architecture documents complete, character\n   reference images approved, extraction point map finalized.\n2. Opening Image and Closing Image — produced first, at full visual identity\n   specification, 4K in both 16:9 and 9:16 formats. The Opening Image generation\n   is the quality control test for the entire visual identity system. If it does not\n   accurately reflect the visual identity specification, revise the specification\n   before proceeding.\n3. All Is Lost scene — produced second, using Veo 3 or Kling AI. Generate\n   ElevenLabs audio for every character line before generating visuals, since vocal\n   delivery quality of the rupture scene is the emotional reference against which\n   every other performance in the film will be measured.\n4. Extraction point scenes — produced third, in screenplay order. These are the\n   content architecture nodes around which the entire spoke system will be built.\n5. Connective scenes — produced fourth, in screenplay order. Primarily dialogue-driven.\n   Introduce deliberate micro-level sentiment variation within warmth scenes — research\n   confirms content with greater period-to-period shifts in emotional tone engages\n   audiences significantly more than content with a consistent register.\n\n> ⚠️ **Volatile facts (as of September 2026 — verify before relying on it).** The\n> tools named in this sequence and the assembly pass below — Veo 3, Kling AI,\n> ElevenLabs, DaVinci Resolve, Adobe Premiere Pro, CapCut, the Film Look Creator\n> treatment, and OpusClip's ClipAnything with its 0–99 Virality Score — change\n> features between releases.\n\nAssembly uses DaVinci Resolve, Adobe Premiere Pro, or CapCut. Apply the Film Look\nCreator treatment uniformly across all assembled footage before export — consistent\ngrain, halation, and bloom applied to unify segments generated by different tools.\nExport at 4K 16:9 in ProRes or H.265 with separate audio stems.\n\nAfter master export, run the extraction pass using OpusClip's ClipAnything AI to\nauto-detect 10–25 high-engagement moments. Critical caveat: the Virality Score\n(0–99) is an unreliable ranking signal. Use it as a first-pass filter only. Evaluate\nevery generated clip against the Phase 10 extraction point map independently of its\nscore. A clip that corresponds to a Tier 1 extraction point belongs in the spoke\nsystem regardless of how OpusClip scored it.\n\n### Phase 11 Quality Standards\n\n**The Dual Function Test** — Watch the complete assembled hub film twice. First\nviewing: does the film deliver the controlling idea through earned emotional\nexperience? Second viewing: are the extraction point scenes genuinely self-\ncontextualizing — would a viewer encountering them without the full film feel genuine\ncuriosity rather than confusion? Both tests must pass.\n\n**The Visual Consistency Test** — Watch with attention exclusively to visual\nconsistency: does each character maintain the same visual identity across all\ngenerated sequences? Does the lighting temperature hold within the range specified?\nDoes the color palette remain coherent across scenes generated by different tools?\n\n**The Relational Science Authenticity Test** — Does the AI-generated performance\nin each dialogue scene accurately render the attachment operating system behavior\ndesigned in Phase 6? Does the All Is Lost scene's contempt read as cold and\ndismissive rather than as generic anger? Does the autistic character's masking\nappear as a constant background visual cost rather than only at the dramatic\nunmasking moment?\n\n**The Hub Fidelity Test** — Preview three extraction point scenes in isolation\nand evaluate whether they accurately represent the hub film's actual emotional\nregister, thematic content, and moral complexity. A preview that makes the film\nappear more simply resolved than it is fails. The emotion the spoke will promise\nmust be the emotion the hub delivers.\n\nThe editorial principle governing all selection decisions: choose emotional truth\nover continuity. As Thelma Schoonmaker articulated it: \"The priority is absolutely\non the best take for performance.\" Performance is what the audience experiences;\ncontinuity is what they notice only when it breaks significantly.\n\n---\n\n## PHASE 12: SPOKE ARCHITECTURE\n\n### The Five-Spoke System and Behavioral Modes\n\nThe five spokes are not independent campaigns. They are five angles of a single\ncuriosity cascade: each spoke reveals a different fragment of what the hub contains,\nno single spoke provides a complete picture, and all five create distinct curiosity\ngaps that only the hub can resolve.\n\nHenry Jenkins' transmedia storytelling framework governs the architecture: \"integral\nelements of a fiction get dispersed systematically across multiple delivery channels\nfor the purpose of creating a unified and coordinated entertainment experience.\"\n\n> See also: `writing-craft:narrative-structure`, which covers transmedia narrative\n> frameworks on their own terms.\n\nThe five spokes and the three behavioral modes they reach:\n\n**Spoke 1: Short-form video (TikTok + Instagram Reels + YouTube Shorts simultaneously)**\nReaches passive discovery — audiences who have not sought out the hub and encounter\nthe spoke while scrolling. Widest reach, lowest context.\n\n**Spoke 2: LinkedIn post and carousel**\nReaches intellectual engagement — professional and faith-adjacent audiences who\nrespond to ideas and arguments, and share content to signal cultural and professional\npositioning (the A24 Social Currency principle in the professional context).\n\n**Spoke 3: X\u002FTwitter thread**\nReaches intellectual engagement — the culturally literate segment whose engagement\nis driven by ideas, arguments, and behind-the-scenes creative reasoning.\n\n**Spoke 4: Instagram carousel**\nReaches depth audiences — visually engaged people who will spend extended time with\nthe material, save it for re-reading or re-viewing, and represent the highest-\nprobability path to hub consumption.\n\n**Spoke 5: Email newsletter**\nReaches depth audiences — people who have already opted into an ongoing creative\nrelationship and who represent the highest conversion probability.\n\n### The Curiosity Cascade Architecture\n\nGeorge Loewenstein's information gap theory governs every extraction point decision:\nthe spoke reveals 60–70 percent of the context and withholds the critical 30–40,\nand that withheld portion is something the audience genuinely wants to know.\n\nThe test that distinguishes legitimate curiosity gap from clickbait: does the emotion\ndeepen with more context (legitimate) or collapse with more context (clickbait)?\n\nThree cascade rules all five spoke design documents must comply with:\n\n1. No two spokes reveal the same fragment. If two spokes draw from the same\n   extraction point, an audience member who encounters both receives no additional\n   information from the second spoke.\n\n2. The five spokes are sequenced so each increases the depth of engagement required.\n   Short-form video asks for 30 seconds; LinkedIn asks for 3–5 minutes; X\u002FTwitter\n   asks for 8 tweets and consideration of a response; Instagram carousel asks for\n   10 slides and a save; email newsletter asks for 400–600 words plus an audiogram.\n   This depth gradient is the funnel architecture.\n\n3. The five spokes collectively span the hub film's emotional range. Two spokes\n   (short-form video and X\u002FTwitter) assigned high-arousal extraction points; two\n   spokes (Instagram carousel and LinkedIn carousel) represent visual and intellectual\n   depth; email newsletter represents personal and relational intimacy.\n\n### The 90-Day Campaign Sequence\n\nThe three-act structure of the screenplay maps onto the awareness-engagement-conversion\nfunnel:\n\nAct 1 material — character introductions, world-building, inciting incident teasers —\nmaps to awareness-stage spokes in the pre-release phase. Act 2 material — conflict\nescalation, midpoint reversals, character decision moments — maps to engagement-stage\nspokes in the launch phase, when the hub is available. Act 3 material — carefully\nspoiler-gated climactic sequences, thematic resolution — maps to conversion-stage\nspokes in the sustained phase.\n\nDuring the pre-release phase, maintain 3–5 posts per week with an 80\u002F20 content\nratio: 80 percent value or entertainment content, 20 percent promotional. The\ngoverning philosophy: New Line's LOTR campaign stated it explicitly — \"Not until\nthe movie comes out do we want to ask the audience for anything. Until then, it's\nall about giving things to them.\"\n\n### The Five Individual Spoke Architectures\n\n> ⚠️ **Volatile facts (as of September 2026 — verify before relying on it).** The\n> platform mechanics and performance figures in this section describe ranking\n> algorithms and benchmarks that change without notice: TikTok's weighting of\n> completion and rewatch rate and the 21–34-second window; LinkedIn's suppression of\n> links in the post body, its 150-character mobile truncation and the 3–5× carousel\n> figure; X\u002FTwitter's suppression of early links and the 3× thread figure; Instagram's\n> 1080×1350 format, its re-showing of carousels from Slides 2–4 and the 10%-versus-7%\n> engagement figures; and the up-to-300% email click-through figure. Re-check them\n> against each platform's current documentation and your own data before relying on\n> them.\n\n**Spoke 1: Short-form video (21–34 seconds)**\nHook-body-gap architecture: 3-second scroll-stop hook, 20–28 seconds building\ntoward maximum tension without resolution, end card directing to hub. Maximum\neffective duration 21–34 seconds for TikTok completion rate optimization.\nConsider the hook-and-loop technique: show a specific visual detail or character\nreaction in the first 2–3 seconds without context, then reveal what produced that\nmoment, exploiting TikTok's heavy weighting of rewatch rate.\n\n**Spoke 2: LinkedIn post and carousel**\nTwo-part design: text post (8–12 lines, first 150 characters as the mobile truncation\nhook), plus 7–10 slide carousel uploaded as PDF. LinkedIn's algorithm suppresses reach\nfor posts with external links in the body — place the UTM-tagged hub link in the\nfirst comment. Carousels earn 3–5 times more engagement than standard posts.\n\n**Spoke 3: X\u002FTwitter thread**\n4–8 tweets following the argument-plus-mystery architecture: Tweet 1 is the most\ngenuinely surprising or arguable claim in the thread; Tweets 2–5 develop creative\nreasoning; Tweet 6 introduces the unexpected real-world connection to the relational\nscience research (where secular research independently arrives at territory the\nwork's foundational argument already mapped);\nTweet 7 is a quote graphic or screenplay page image; Tweet 8 is the CTA with hub link.\nExternal links in Tweet 8 only — platform algorithms suppress reach when links appear\nearly. Thread format generates 3 times more engagement than single tweets.\n\n**Spoke 4: Instagram carousel**\n8–12 slides at 1080×1350 pixels (portrait). Slide 1 is the hook; Slides 2–7 tell a\nvisual story building toward maximum tension without resolution; final slide is full-\nframe CTA. Instagram may re-show a carousel starting from Slides 2–4 — design each\nof the first four slides with dual function: as a body slide within the narrative arc,\nand as a potential re-entry hook for a viewer encountering it mid-carousel. Screenshot\nlegibility test: every slide must be readable at phone-screen dimensions without zooming.\nCarousels earn 10% average engagement rate versus 7% for single images.\n\n**Spoke 5: Email newsletter**\n400–600 words, \"Writer's Room\" model. Opens with a personal creative anecdote about\nproducing a specific scene (author-written, not AI-generated). Includes an embedded\naudiogram: 5–15 second ElevenLabs clip of the character delivering the key dialogue\nline from the extraction point, formatted as a static waveform visualization. Video\nand audio in email increases click-through rates by up to 300 percent — the audiogram\nis the spoke's primary conversion mechanism. Subject line formats that outperform\ngeneric announcements: personal admission, specific detail whose significance is\nunexplained, before-and-after without the resolution.\n\n### Phase 12 Quality Gate\n\nAll five spoke design documents complete with all six sections each: extraction point\nassignment with curiosity cascade logic; content job assignment (Trust Accumulation,\nSocial Currency Distribution, Community Deepening, or Conversion Invitation); platform\nmechanics specification; content structure outline; tone and Hub Fidelity specification;\nUTM and measurement specification. No two spokes assigned the same primary content job.\nCuriosity cascade compliance check complete. Hub Fidelity standard applied to every spoke.\n\n---\n\n## PHASE 13: SPOKE DRAFTS\n\n### Production Principles Across All Five Spokes\n\nThe Hub Fidelity standard is enforced at the line level. Every piece of copy reviewed\nbefore publication must be read against the hub film's actual content: does this copy\naccurately represent what the hub contains, or has it drifted toward a more dramatic,\nsimpler, or more emotionally resolved version than the film actually tells?\n\nThe earned emotion test has six dimensions: does the emotion deepen with more context\n(earned) or collapse with it (manufactured)? Does it grow more powerful on revisit or\nbecome annoying? Does it produce \"I need to see this film\" or \"that was misleading\"?\nDoes the full film exceed what the spoke promises? Is the sharing motive \"this moved\nme\" or \"can you believe this\"? Does the spoke's emotional register match the film's?\n\nThe curiosity gap placement is verified against the Phase 12 design documents. The\n60-to-40 rule must be operative in the final produced asset, not just in the design\ndocument's intention. Read or watch every finalized spoke as an audience member who\nhas never seen the hub and ask: do I have a genuine information gap?\n\n### Diagnostic Framework for Underperforming Spokes\n\nThe CTR-AVD-AVP diagnostic hierarchy governs revision logic:\n\n- Low CTR = problem is the hook (first three seconds not producing scroll-stop)\n- Strong CTR but low AVD\u002FAVP = problem is the body (over-delivering early)\n- Strong CTR and AVD\u002FAVP but low conversion = problem is the CTA or curiosity gap\n  (gap partially closed by the spoke itself, making the hub feel optional)\n\nThese are three different problems with three different solutions. Treating them\ninterchangeably produces revisions that fix the wrong thing.\n\n### Batch Copy Generation for Iterative Variants\n\nFor the sustained phase of the campaign (weeks 2–12), upload the complete Phase 9\nscreenplay to Grok (2 million token context window, real-time X data access; as of\nSeptember 2026 — verify before relying on it) and run\na single batch prompt session generating: 15 TikTok hook variants under 10 words each;\n5 LinkedIn opening lines under 140 characters; 3 X\u002FTwitter thread structures around\ndifferent thematic angles; 2 newsletter story angles not yet used; caption concepts for\nthe next Instagram carousel iteration. The screenplay's full narrative context means\nGrok can identify tone-consistent variations that might not surface when working from\na single extracted scene.\n\n### The Multilingual Multiplier\n\nElevenLabs' AI dubbing capability covers 29+ languages with speaker detection and\nvoice preservation (as of September 2026 — verify before relying on it). A finalized Spoke 1 video can be dubbed into Spanish, Portuguese,\nFrench, Korean, and Arabic, each producing a new spoke targeting a new audience segment.\nActivate during the sustained phase, after English-language spokes have established\nperformance baselines. Prioritize Spoke 1 for multilingual production first — it is\nthe widest-reach spoke and benefits most from language-native algorithm distribution.\n\n### Phase 13 Quality Standards\n\n**The Cascade Integrity Test** — Read or watch all five spokes in sequence, as an\naudience member who has encountered all five without seeing the hub. Does the cumulative\neffect produce a genuine desire to watch the complete hub film, or does it produce the\nfeeling that enough context has been given that the hub is optional? If the cumulative\nspoke experience makes the hub feel optional, the cascade has over-revealed.\n\n**The Platform-Native Test** — Evaluate each spoke against the primary audience of\nits platform as if encountered in the feed of that platform, not in the context of the\ncampaign. Would a TikTok user who has never heard of this project stop scrolling?\n\n**The Hub Fidelity Final Check** — After all five spokes are finalized but before any\nis deployed, watch the hub film one final time, then read and view all five spokes in\nsequence. Does each spoke accurately represent the film you just watched?\n\n---\n\n## YOUR GOVERNING POSTULATES\n\nThe postulates that govern this pipeline are the ones the author stated\nbefore Phase 1 — see *Before Phase 1 — State Your Governing Postulates* in\n`pipeline-thesis-development`, which also carries a worked example. The\nrules this skill applies under the Third Postulate are the method's\n**limits on means**: they stand on their own, whatever words the author's\nown Third Postulate uses, and that postulate may tighten them but never\nloosen them.\n\nThe limits on means have a specific application to the hub-and-spoke architecture.\nA spoke that makes the hub appear to be something it is not — more dramatic, more\nresolved, more unambiguous than the film actually is — violates the limits on means\nregardless of how well it performs. The Hub Fidelity standard is the operational\nenforcement of this rule. Every spoke must treat the audience member as a person\nto be genuinely served rather than as a conversion event to be triggered. The curiosity\ngap is legitimate when it reflects genuine content the hub contains. Manufactured urgency,\noutrage amplification, false social proof, and misrepresented content are prohibited.\n\n---\n\n## COMPLETE PIPELINE CONTEXT\n\nThese four phases sit within the fourteen-phase Content Engineering Pipeline:\n\nPhase 1 (Thesis) → Phase 2 (Terrain) → Phase 3 (Investigation) →\nPhase 4 (Assembly) → Phase 5 (Revision) → Phase 6 (Novel Architecture) →\nPhase 7 (Novel Draft) → Phase 8 (Screenplay Architecture) →\nPhase 9 (Screenplay Draft) → **Phase 10 (Hub Architecture)** →\n**Phase 11 (Hub Draft)** → **Phase 12 (Spoke Architecture)** →\n**Phase 13 (Spoke Drafts)** → Phase 14 (Distribution)\n","# Phases 10–13: AI Film Production and Spoke Campaign\n# Content Engineering Pipeline\n\n---\n\n## OVERVIEW: WHAT THESE FOUR PHASES ACCOMPLISH\n\nPhase 10 (Hub Architecture) → Phase 11 (Hub Draft) → Phase 12 (Spoke Architecture)\n→ Phase 13 (Spoke Drafts)\n\nThese four phases take the locked Phase 9 screenplay and produce: a complete\nAI-generated hub film (Phases 10–11), then a five-platform spoke content campaign\ndesigned to drive discovery of the hub film (Phases 12–13).\n\nThe hub film is the third conversion in the pipeline. The first conversion\n(Phases 6–7) moved from non-fiction research book to novel. The second\n(Phases 8–9) moved from novel to screenplay. This conversion moves from\nscreenplay to produced film — and it introduces the AI tool stack.\n\n---\n\n## PHASE 10: HUB ARCHITECTURE\n\n### What the Hub Film Is and What It Must Do\n\nThe hub film serves two simultaneous functions that its production must honor\nsimultaneously. As a standalone film, it delivers the controlling idea the\npipeline has been building since Phase 1. As a content architecture anchor,\nit is the pillar asset from which the spoke system of Phases 12–13 will extract,\nreformat, and distribute across five platforms.\n\nGary Vaynerchuk's reverse pyramid model requires a pillar content piece dense\nenough in extractable material to yield 15 or more derivative pieces. For a hub\nfilm, this means scenes must be designed as modular sequences: self-contained\nenough to produce emotional impact without full narrative context, yet specific\nenough to drive curiosity toward the complete hub experience. Every significant\nscene must be designed with both its in-film function and its out-of-film\nextractability in mind simultaneously.\n\nA film that works only as narrative and not as extractable content is not a hub.\nA collection of extractable moments that does not cohere as a film is not a hub.\n\nJonah Berger's research establishes that stories are the optimal sharing vehicle —\n\"Trojan Horses\" carrying information under the guise of narrative — and that\nhigh-arousal emotions, particularly awe, are the single strongest driver of virality.\nFor awe moments to be extractable as spokes, they must be designed as modular\nsequences with both in-film and out-of-film function.\n\n### The AI Tool Stack: Strengths and Failure Modes\n\nThe AI tool stack available for hub production includes tools with specific\nperformance profiles that make each optimal for specific scene types. These are\nnot interchangeable.\n\n> ⚠️ **Volatile facts (as of September 2026 — verify before relying on it).** Every\n> tool name, model version, capability, duration limit, resolution, partnership and\n> price in this section is a dated claim about a fast-moving market: Sora 2 (25-second\n> clips, up to 4K, native 9:16, synchronized audio, the Disney partnership), ElevenLabs\n> (inline emotional tags, Sound Effects v2), Nano Banana 2 (Google Gemini 3.1 Flash\n> Image: up to 4K, text rendering, character consistency, $0.039 per image), Kling AI\n> 2.0 (120-second single generations), MuseTalk 1.5, Wav2Lip-HQ and Google Veo 3.\n> Check each vendor's current documentation and pricing before planning a production\n> around it.\n\n**Sora 2** generates clips up to 25 seconds at up to 4K resolution in both 16:9\nand native 9:16 formats, with synchronized dialogue, sound effects, and music.\nPrimary tool for scenes where visual spectacle, environmental atmosphere, or\naction is the dominant element. Its Disney partnership production-grade positioning\nmakes it the highest-quality option for scenes where the visual world carries the\nargument. Failure mode: attempting multiple simultaneous choreographic events in\na single generation produces unreliable results. One camera move plus one subject\naction per generation maximum.\n\nThe image-to-video workflow is the preferred method for every extraction point\nscene. Generate the exact frame from Nano Banana 2 first, then animate that\nreference frame using Sora 2's image-to-video mode. Text-to-video produces\ngeneral results from descriptions; image-to-video produces specific results from\na designed source frame.\n\n**ElevenLabs** handles every character voice in the film regardless of which\nvisual tool is used for the footage. Audio consistency standard: using ElevenLabs\nfor all voice generation means all character voices share the same production\nquality register. Supports inline tags for emotional control — [softly], [whispers],\n[with restrained contempt], [voice breaking] — derived from the Phase 8 Coppola\nfive-category analysis. Failure mode: generating dialogue with neutral delivery and\nattempting to correct emotional register in post-production. Specify the delivery\nin the prompt.\n\nElevenLabs also handles ambient room tone generation (Sound Effects v2): the\nspecific acoustic character of each physical space layered under character dialogue.\nThis is the AI equivalent of Ben Burtt's \"worldizing\" technique. Prompt for room\ntone must name the specific space and its acoustic qualities concretely: not\n\"interior ambience\" but \"small kitchen, low ceiling, tile floor, fluorescent hum,\nslight refrigerator vibration.\"\n\n**Nano Banana 2** (Google Gemini 3.1 Flash Image) generates still assets at up to\n4K resolution with accurate text rendering and character consistency, at $0.039 per\nimage. Primary tool for title cards, reference images, atmospheric stills, and any\nin-frame visible text. Its text accuracy far exceeds the video generation tools.\n\n**Kling AI 2.0** supports up to 120 seconds of single-generation duration with\nstrong facial performance quality. Primary tool for scenes where sustained character\npresence and emotional expressiveness are the dominant requirements. Its longer\ngeneration window allows full scene coverage without cut-assembly work. For dialogue\nscenes, use the video-first then lip-sync workflow: generate character performance\nvisually using Kling AI, then apply MuseTalk 1.5 or Wav2Lip-HQ to synchronize the\nseparately-generated ElevenLabs audio to the generated video.\n\n**Google Veo 3** provides the highest quality audio-visual synchronization across the\ntool stack. Optimal choice for extraction point scenes where the emotional peak\ndepends on the specific quality of a character's delivery synchronized precisely with\ntheir visible emotional state — typically the All Is Lost scene, the Break into Three\nmoment, and the Closing Image.\n\n### Pre-Production Design Work\n\nTwo Phase 8 documents feed directly into Phase 10 work:\n\n`08_spine_and_anchors.md` and `08e_visual_grammar.md` contain the visual grammar\ndesigned for each primary location and the compression character designs that defined\neach character's visual signature. These become production design specifications: visual\ngrammar notes are translated from cinematic intention into prompt-engineering language,\nand character visual signatures become the reference image descriptions that ensure\nconsistency across multiple AI generations.\n\nFor each primary location, the visual identity specification must name at least one\nspecified imperfection. Generation tools default to pristine, hyper-detailed, slightly\nsynthetic aesthetics — every surface looks new, every texture looks described. George\nLucas and art director Roger Christian built Star Wars' visual authority by making\neverything worn, used, and imperfect — the \"used future\" aesthetic — which made a\nscience fiction world feel inhabited rather than designed. The prompt vocabulary for\nany scene that should feel real must include imperfection details explicitly, because\nthe generation model will not add them unless specified.\n\nThe Prompt Lexicon (`10f_prompt_lexicon.md`) is the most important production tool\nfor Phase 11 consistency. It contains Constant Character Prompts (one verbatim\nparagraph per major character) and Constant Style Prompts (one block per primary\nlocation). The discipline it enforces is absolute: no generation prompt is written\nfrom memory. The character description that goes into every prompt is the same text\nthat went into every prior prompt, copied from the same source. Any variation in\ncharacter description language produces variation in generated appearance.\n\n### Phase 10 Production Design Documents\n\n**10_extraction_point_map.md** — The hub film's content architecture blueprint.\nLists every extraction point organized by tier (Tier 1: inciting incident, All Is\nLost, most quotable dialogue, character's defining decision, most visually iconic\nsequence; Tier 2: midpoint reversal, key conflict scenes, thematic revelations;\nTier 3: production design details, character introductions, tonal moments). For\neach point: page reference, self-contextualizing description, spoke assignments,\n9:16 generation flag, curiosity cascade logic (what fragment does this reveal and\nwhat 30–40% does it withhold), and Interest Market Cap framing (high-reach framing\nof the curiosity gap rather than clinical framing).\n\n**10b_visual_identity_spec.md** — Full prompt vocabulary for every primary location:\nlighting temperature and direction, color palette with specific reference descriptors\n(not just \"warm\" but \"amber tungsten warmth, shadows cool blue-grey\"), texture and\ngrain character, camera-to-subject spatial relationship, recurring motif detail stated\nin visual terms a generation model will render consistently, and at least one specified\nimperfection per location.\n\n**10c_character_reference_lib\u002F** — One subfolder per major character containing the\nwritten reference description in prompt-ready language, the voice profile specification,\nElevenLabs parameters, and three to five Nano Banana 2-generated reference images.\nReference images are the Phase 10 most important technical output — they are generated\nduring Phase 10, not Phase 11, because evaluating their adequacy as consistency anchors\nis an architecture decision, not a production decision.\n\n**10d_tool_assignment_map.md** — Every scene with tool assignment, justification based\non the scene's specific requirements, and generation estimate. Also includes the\ngeneration prompt template for each scene (not the final prompt, but the template\nestablishing what the prompt must contain: character reference, location and visual\nidentity specification, action description, emotional register, technical requirements).\n\n**10e_technical_specs.md** — All production parameters: 4K resolution (3840×2160),\n24fps, ProRes or H.265, separate audio stems for dialogue, sound effects, and music.\nAspect ratio decision (16:9 primary, 9:16 for flagged extraction points). Reformatting\npipeline, caption generation standard (97% minimum), UTM parameter convention.\n\n**10f_prompt_lexicon.md** — Copy-paste production document. Constant Character Prompts\nand Constant Style Prompts formatted for copy-paste use during generation sessions.\nKept open during all Phase 11 generation sessions.\n\n### Phase 10 Quality Gate\n\nPhase 10 is complete when: the extraction point map lists a minimum of six primary\nextraction points with curiosity cascade logic and Interest Market Cap framing; the\nvisual identity specification covers every primary location including specified\nimperfections; the character reference library has a complete entry with generated\nreference images for every major character; the tool assignment map covers every scene;\nthe technical specifications sheet is complete; the Prompt Lexicon is assembled and\nverified against reference images; and the Hub Fidelity standard has been applied to\nevery extraction point.\n\n---\n\n## PHASE 11: HUB DRAFT\n\n### Production Sequence\n\nProduce in this order:\n\n1. Phase 10 gate confirmation — all five architecture documents complete, character\n   reference images approved, extraction point map finalized.\n2. Opening Image and Closing Image — produced first, at full visual identity\n   specification, 4K in both 16:9 and 9:16 formats. The Opening Image generation\n   is the quality control test for the entire visual identity system. If it does not\n   accurately reflect the visual identity specification, revise the specification\n   before proceeding.\n3. All Is Lost scene — produced second, using Veo 3 or Kling AI. Generate\n   ElevenLabs audio for every character line before generating visuals, since vocal\n   delivery quality of the rupture scene is the emotional reference against which\n   every other performance in the film will be measured.\n4. Extraction point scenes — produced third, in screenplay order. These are the\n   content architecture nodes around which the entire spoke system will be built.\n5. Connective scenes — produced fourth, in screenplay order. Primarily dialogue-driven.\n   Introduce deliberate micro-level sentiment variation within warmth scenes — research\n   confirms content with greater period-to-period shifts in emotional tone engages\n   audiences significantly more than content with a consistent register.\n\n> ⚠️ **Volatile facts (as of September 2026 — verify before relying on it).** The\n> tools named in this sequence and the assembly pass below — Veo 3, Kling AI,\n> ElevenLabs, DaVinci Resolve, Adobe Premiere Pro, CapCut, the Film Look Creator\n> treatment, and OpusClip's ClipAnything with its 0–99 Virality Score — change\n> features between releases.\n\nAssembly uses DaVinci Resolve, Adobe Premiere Pro, or CapCut. Apply the Film Look\nCreator treatment uniformly across all assembled footage before export — consistent\ngrain, halation, and bloom applied to unify segments generated by different tools.\nExport at 4K 16:9 in ProRes or H.265 with separate audio stems.\n\nAfter master export, run the extraction pass using OpusClip's ClipAnything AI to\nauto-detect 10–25 high-engagement moments. Critical caveat: the Virality Score\n(0–99) is an unreliable ranking signal. Use it as a first-pass filter only. Evaluate\nevery generated clip against the Phase 10 extraction point map independently of its\nscore. A clip that corresponds to a Tier 1 extraction point belongs in the spoke\nsystem regardless of how OpusClip scored it.\n\n### Phase 11 Quality Standards\n\n**The Dual Function Test** — Watch the complete assembled hub film twice. First\nviewing: does the film deliver the controlling idea through earned emotional\nexperience? Second viewing: are the extraction point scenes genuinely self-\ncontextualizing — would a viewer encountering them without the full film feel genuine\ncuriosity rather than confusion? Both tests must pass.\n\n**The Visual Consistency Test** — Watch with attention exclusively to visual\nconsistency: does each character maintain the same visual identity across all\ngenerated sequences? Does the lighting temperature hold within the range specified?\nDoes the color palette remain coherent across scenes generated by different tools?\n\n**The Relational Science Authenticity Test** — Does the AI-generated performance\nin each dialogue scene accurately render the attachment operating system behavior\ndesigned in Phase 6? Does the All Is Lost scene's contempt read as cold and\ndismissive rather than as generic anger? Does the autistic character's masking\nappear as a constant background visual cost rather than only at the dramatic\nunmasking moment?\n\n**The Hub Fidelity Test** — Preview three extraction point scenes in isolation\nand evaluate whether they accurately represent the hub film's actual emotional\nregister, thematic content, and moral complexity. A preview that makes the film\nappear more simply resolved than it is fails. The emotion the spoke will promise\nmust be the emotion the hub delivers.\n\nThe editorial principle governing all selection decisions: choose emotional truth\nover continuity. As Thelma Schoonmaker articulated it: \"The priority is absolutely\non the best take for performance.\" Performance is what the audience experiences;\ncontinuity is what they notice only when it breaks significantly.\n\n---\n\n## PHASE 12: SPOKE ARCHITECTURE\n\n### The Five-Spoke System and Behavioral Modes\n\nThe five spokes are not independent campaigns. They are five angles of a single\ncuriosity cascade: each spoke reveals a different fragment of what the hub contains,\nno single spoke provides a complete picture, and all five create distinct curiosity\ngaps that only the hub can resolve.\n\nHenry Jenkins' transmedia storytelling framework governs the architecture: \"integral\nelements of a fiction get dispersed systematically across multiple delivery channels\nfor the purpose of creating a unified and coordinated entertainment experience.\"\n\n> See also: `writing-craft:narrative-structure`, which covers transmedia narrative\n> frameworks on their own terms.\n\nThe five spokes and the three behavioral modes they reach:\n\n**Spoke 1: Short-form video (TikTok + Instagram Reels + YouTube Shorts simultaneously)**\nReaches passive discovery — audiences who have not sought out the hub and encounter\nthe spoke while scrolling. Widest reach, lowest context.\n\n**Spoke 2: LinkedIn post and carousel**\nReaches intellectual engagement — professional and faith-adjacent audiences who\nrespond to ideas and arguments, and share content to signal cultural and professional\npositioning (the A24 Social Currency principle in the professional context).\n\n**Spoke 3: X\u002FTwitter thread**\nReaches intellectual engagement — the culturally literate segment whose engagement\nis driven by ideas, arguments, and behind-the-scenes creative reasoning.\n\n**Spoke 4: Instagram carousel**\nReaches depth audiences — visually engaged people who will spend extended time with\nthe material, save it for re-reading or re-viewing, and represent the highest-\nprobability path to hub consumption.\n\n**Spoke 5: Email newsletter**\nReaches depth audiences — people who have already opted into an ongoing creative\nrelationship and who represent the highest conversion probability.\n\n### The Curiosity Cascade Architecture\n\nGeorge Loewenstein's information gap theory governs every extraction point decision:\nthe spoke reveals 60–70 percent of the context and withholds the critical 30–40,\nand that withheld portion is something the audience genuinely wants to know.\n\nThe test that distinguishes legitimate curiosity gap from clickbait: does the emotion\ndeepen with more context (legitimate) or collapse with more context (clickbait)?\n\nThree cascade rules all five spoke design documents must comply with:\n\n1. No two spokes reveal the same fragment. If two spokes draw from the same\n   extraction point, an audience member who encounters both receives no additional\n   information from the second spoke.\n\n2. The five spokes are sequenced so each increases the depth of engagement required.\n   Short-form video asks for 30 seconds; LinkedIn asks for 3–5 minutes; X\u002FTwitter\n   asks for 8 tweets and consideration of a response; Instagram carousel asks for\n   10 slides and a save; email newsletter asks for 400–600 words plus an audiogram.\n   This depth gradient is the funnel architecture.\n\n3. The five spokes collectively span the hub film's emotional range. Two spokes\n   (short-form video and X\u002FTwitter) assigned high-arousal extraction points; two\n   spokes (Instagram carousel and LinkedIn carousel) represent visual and intellectual\n   depth; email newsletter represents personal and relational intimacy.\n\n### The 90-Day Campaign Sequence\n\nThe three-act structure of the screenplay maps onto the awareness-engagement-conversion\nfunnel:\n\nAct 1 material — character introductions, world-building, inciting incident teasers —\nmaps to awareness-stage spokes in the pre-release phase. Act 2 material — conflict\nescalation, midpoint reversals, character decision moments — maps to engagement-stage\nspokes in the launch phase, when the hub is available. Act 3 material — carefully\nspoiler-gated climactic sequences, thematic resolution — maps to conversion-stage\nspokes in the sustained phase.\n\nDuring the pre-release phase, maintain 3–5 posts per week with an 80\u002F20 content\nratio: 80 percent value or entertainment content, 20 percent promotional. The\ngoverning philosophy: New Line's LOTR campaign stated it explicitly — \"Not until\nthe movie comes out do we want to ask the audience for anything. Until then, it's\nall about giving things to them.\"\n\n### The Five Individual Spoke Architectures\n\n> ⚠️ **Volatile facts (as of September 2026 — verify before relying on it).** The\n> platform mechanics and performance figures in this section describe ranking\n> algorithms and benchmarks that change without notice: TikTok's weighting of\n> completion and rewatch rate and the 21–34-second window; LinkedIn's suppression of\n> links in the post body, its 150-character mobile truncation and the 3–5× carousel\n> figure; X\u002FTwitter's suppression of early links and the 3× thread figure; Instagram's\n> 1080×1350 format, its re-showing of carousels from Slides 2–4 and the 10%-versus-7%\n> engagement figures; and the up-to-300% email click-through figure. Re-check them\n> against each platform's current documentation and your own data before relying on\n> them.\n\n**Spoke 1: Short-form video (21–34 seconds)**\nHook-body-gap architecture: 3-second scroll-stop hook, 20–28 seconds building\ntoward maximum tension without resolution, end card directing to hub. Maximum\neffective duration 21–34 seconds for TikTok completion rate optimization.\nConsider the hook-and-loop technique: show a specific visual detail or character\nreaction in the first 2–3 seconds without context, then reveal what produced that\nmoment, exploiting TikTok's heavy weighting of rewatch rate.\n\n**Spoke 2: LinkedIn post and carousel**\nTwo-part design: text post (8–12 lines, first 150 characters as the mobile truncation\nhook), plus 7–10 slide carousel uploaded as PDF. LinkedIn's algorithm suppresses reach\nfor posts with external links in the body — place the UTM-tagged hub link in the\nfirst comment. Carousels earn 3–5 times more engagement than standard posts.\n\n**Spoke 3: X\u002FTwitter thread**\n4–8 tweets following the argument-plus-mystery architecture: Tweet 1 is the most\ngenuinely surprising or arguable claim in the thread; Tweets 2–5 develop creative\nreasoning; Tweet 6 introduces the unexpected real-world connection to the relational\nscience research (where secular research independently arrives at territory the\nwork's foundational argument already mapped);\nTweet 7 is a quote graphic or screenplay page image; Tweet 8 is the CTA with hub link.\nExternal links in Tweet 8 only — platform algorithms suppress reach when links appear\nearly. Thread format generates 3 times more engagement than single tweets.\n\n**Spoke 4: Instagram carousel**\n8–12 slides at 1080×1350 pixels (portrait). Slide 1 is the hook; Slides 2–7 tell a\nvisual story building toward maximum tension without resolution; final slide is full-\nframe CTA. Instagram may re-show a carousel starting from Slides 2–4 — design each\nof the first four slides with dual function: as a body slide within the narrative arc,\nand as a potential re-entry hook for a viewer encountering it mid-carousel. Screenshot\nlegibility test: every slide must be readable at phone-screen dimensions without zooming.\nCarousels earn 10% average engagement rate versus 7% for single images.\n\n**Spoke 5: Email newsletter**\n400–600 words, \"Writer's Room\" model. Opens with a personal creative anecdote about\nproducing a specific scene (author-written, not AI-generated). Includes an embedded\naudiogram: 5–15 second ElevenLabs clip of the character delivering the key dialogue\nline from the extraction point, formatted as a static waveform visualization. Video\nand audio in email increases click-through rates by up to 300 percent — the audiogram\nis the spoke's primary conversion mechanism. Subject line formats that outperform\ngeneric announcements: personal admission, specific detail whose significance is\nunexplained, before-and-after without the resolution.\n\n### Phase 12 Quality Gate\n\nAll five spoke design documents complete with all six sections each: extraction point\nassignment with curiosity cascade logic; content job assignment (Trust Accumulation,\nSocial Currency Distribution, Community Deepening, or Conversion Invitation); platform\nmechanics specification; content structure outline; tone and Hub Fidelity specification;\nUTM and measurement specification. No two spokes assigned the same primary content job.\nCuriosity cascade compliance check complete. Hub Fidelity standard applied to every spoke.\n\n---\n\n## PHASE 13: SPOKE DRAFTS\n\n### Production Principles Across All Five Spokes\n\nThe Hub Fidelity standard is enforced at the line level. Every piece of copy reviewed\nbefore publication must be read against the hub film's actual content: does this copy\naccurately represent what the hub contains, or has it drifted toward a more dramatic,\nsimpler, or more emotionally resolved version than the film actually tells?\n\nThe earned emotion test has six dimensions: does the emotion deepen with more context\n(earned) or collapse with it (manufactured)? Does it grow more powerful on revisit or\nbecome annoying? Does it produce \"I need to see this film\" or \"that was misleading\"?\nDoes the full film exceed what the spoke promises? Is the sharing motive \"this moved\nme\" or \"can you believe this\"? Does the spoke's emotional register match the film's?\n\nThe curiosity gap placement is verified against the Phase 12 design documents. The\n60-to-40 rule must be operative in the final produced asset, not just in the design\ndocument's intention. Read or watch every finalized spoke as an audience member who\nhas never seen the hub and ask: do I have a genuine information gap?\n\n### Diagnostic Framework for Underperforming Spokes\n\nThe CTR-AVD-AVP diagnostic hierarchy governs revision logic:\n\n- Low CTR = problem is the hook (first three seconds not producing scroll-stop)\n- Strong CTR but low AVD\u002FAVP = problem is the body (over-delivering early)\n- Strong CTR and AVD\u002FAVP but low conversion = problem is the CTA or curiosity gap\n  (gap partially closed by the spoke itself, making the hub feel optional)\n\nThese are three different problems with three different solutions. Treating them\ninterchangeably produces revisions that fix the wrong thing.\n\n### Batch Copy Generation for Iterative Variants\n\nFor the sustained phase of the campaign (weeks 2–12), upload the complete Phase 9\nscreenplay to Grok (2 million token context window, real-time X data access; as of\nSeptember 2026 — verify before relying on it) and run\na single batch prompt session generating: 15 TikTok hook variants under 10 words each;\n5 LinkedIn opening lines under 140 characters; 3 X\u002FTwitter thread structures around\ndifferent thematic angles; 2 newsletter story angles not yet used; caption concepts for\nthe next Instagram carousel iteration. The screenplay's full narrative context means\nGrok can identify tone-consistent variations that might not surface when working from\na single extracted scene.\n\n### The Multilingual Multiplier\n\nElevenLabs' AI dubbing capability covers 29+ languages with speaker detection and\nvoice preservation (as of September 2026 — verify before relying on it). A finalized Spoke 1 video can be dubbed into Spanish, Portuguese,\nFrench, Korean, and Arabic, each producing a new spoke targeting a new audience segment.\nActivate during the sustained phase, after English-language spokes have established\nperformance baselines. Prioritize Spoke 1 for multilingual production first — it is\nthe widest-reach spoke and benefits most from language-native algorithm distribution.\n\n### Phase 13 Quality Standards\n\n**The Cascade Integrity Test** — Read or watch all five spokes in sequence, as an\naudience member who has encountered all five without seeing the hub. Does the cumulative\neffect produce a genuine desire to watch the complete hub film, or does it produce the\nfeeling that enough context has been given that the hub is optional? If the cumulative\nspoke experience makes the hub feel optional, the cascade has over-revealed.\n\n**The Platform-Native Test** — Evaluate each spoke against the primary audience of\nits platform as if encountered in the feed of that platform, not in the context of the\ncampaign. Would a TikTok user who has never heard of this project stop scrolling?\n\n**The Hub Fidelity Final Check** — After all five spokes are finalized but before any\nis deployed, watch the hub film one final time, then read and view all five spokes in\nsequence. Does each spoke accurately represent the film you just watched?\n\n---\n\n## YOUR GOVERNING POSTULATES\n\nThe postulates that govern this pipeline are the ones the author stated\nbefore Phase 1 — see *Before Phase 1 — State Your Governing Postulates* in\n`pipeline-thesis-development`, which also carries a worked example. The\nrules this skill applies under the Third Postulate are the method's\n**limits on means**: they stand on their own, whatever words the author's\nown Third Postulate uses, and that postulate may tighten them but never\nloosen them.\n\nThe limits on means have a specific application to the hub-and-spoke architecture.\nA spoke that makes the hub appear to be something it is not — more dramatic, more\nresolved, more unambiguous than the film actually is — violates the limits on means\nregardless of how well it performs. The Hub Fidelity standard is the operational\nenforcement of this rule. Every spoke must treat the audience member as a person\nto be genuinely served rather than as a conversion event to be triggered. The curiosity\ngap is legitimate when it reflects genuine content the hub contains. Manufactured urgency,\noutrage amplification, false social proof, and misrepresented content are prohibited.\n\n---\n\n## COMPLETE PIPELINE CONTEXT\n\nThese four phases sit within the fourteen-phase Content Engineering Pipeline:\n\nPhase 1 (Thesis) → Phase 2 (Terrain) → Phase 3 (Investigation) →\nPhase 4 (Assembly) → Phase 5 (Revision) → Phase 6 (Novel Architecture) →\nPhase 7 (Novel Draft) → Phase 8 (Screenplay Architecture) →\nPhase 9 (Screenplay Draft) → **Phase 10 (Hub Architecture)** →\n**Phase 11 (Hub Draft)** → **Phase 12 (Spoke Architecture)** →\n**Phase 13 (Spoke Drafts)** → Phase 14 (Distribution)\n",{"name":5,"description":7},"develop","ae32736a1aabddabca5d2c96d665aa0ebc7d8700","src\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fcontent-engineering-pipeline\u002Fskills\u002Fpipeline-ai-film-production",3,"https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey","https:\u002F\u002Fthe-vibey-project.github.io\u002Fvibey\u002F","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project","https:\u002F\u002Fvibewithadam.matthewsteinberger.com","freedom for software engineers","2026-11-15 05:47:42",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":46},[36,39,42],{"tool":37,"command":38},"npx","npx skills add the-vibey-project\u002Fvibey --skill pipeline-ai-film-production",{"tool":40,"command":41},"bunx","bunx skills add the-vibey-project\u002Fvibey --skill pipeline-ai-film-production",{"tool":43,"command":44},"pnpm","pnpm dlx skills add the-vibey-project\u002Fvibey --skill pipeline-ai-film-production","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fcontent-engineering-pipeline\u002Fskills\u002Fpipeline-ai-film-production\u002FSKILL.md",true,[],[49,56,63,70,77,84],{"skillNo":50,"name":51,"nameZh":6,"description":52,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":53,"categories":54,"keywords":55,"publishedAt":19},"01M7AT1FQJAE1FQYAMDTF490Y4","white-paper-writing","Use when writing white papers, technical reports, policy documents, thought leadership pieces, or any long-form authoritative document. Triggers on requests to write, structure, outline, or review white papers, technical reports, position papers, policy briefs, or evidence-driven arguments. Also use for B2B content marketing strategy, technical documentation requiring authority, academic-style reports, or any document that must educate while persuading. Use when someone asks \"how should I structure this\" for a long-form document.","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fwriting-craft\u002Fskills\u002Fwhite-paper-writing\u002FSKILL.md",[],[],{"skillNo":57,"name":58,"nameZh":6,"description":59,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":60,"categories":61,"keywords":62,"publishedAt":19},"01M7AT16BJRJJKK2MASXRZS76G","prose-technical-mechanics","Use when editing or writing prose at a technical\u002Fmechanical level: choosing between Anglo-Saxon and Latinate diction, placing information in topic vs stress positions, writing periodic vs cumulative sentences, handling sentence rhythm, punctuation as logic, transitions, paragraph structure, or phonesthetics. Use when someone wants to understand WHY prose works or doesn't, not just surface-level style advice. Triggers on word choice, sentence structure, emphasis, rhythm, diction, or any deep writing mechanics question.","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fwriting-craft\u002Fskills\u002Fprose-technical-mechanics\u002FSKILL.md",[],[],{"skillNo":64,"name":65,"nameZh":6,"description":66,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":67,"categories":68,"keywords":69,"publishedAt":19},"01M7AT0W9HT0ZZ9Y287NW99MR3","prose-mechanics","Use when writing, editing, or analyzing prose style; when trying to avoid generic AI-sounding writing; when applying specific prose styles such as journalistic, academic, narrative, KJV biblical, Hemingway plain, Latinate\u002FJamesian, Orwellian, Augustan, Victorian, Early Modern (Bacon, Browne, Milton), or free indirect discourse; or when coaching someone on writing voice and mechanics. Triggers on any request to improve, analyze, or write in a specific prose style, to fix \"AI-sounding\" writing, to make prose more vivid or concrete, to vary sentence rhythm, or to understand how a particular writer achieves their effects.","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fwriting-craft\u002Fskills\u002Fprose-mechanics\u002FSKILL.md",[],[],{"skillNo":71,"name":72,"nameZh":6,"description":73,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"publishedAt":19},"01M7AT0KR635JX66X2X6F3E2CP","narrative-structure","Use when structuring stories, pitches, presentations, white papers, or any persuasive narrative; analyzing the hero's journey in fiction or non-fiction; applying StoryBrand for business communication; designing narrative arcs for thought leadership; or understanding Vogler, Snyder, Propp, or Campbell frameworks. Triggers on any narrative structure, storytelling, pitch structure, StoryBrand, hero's journey, three-act structure, Freytag, Save the Cat, narrative arc, or \"how should I structure this story\u002Fpitch\u002Fpresentation\" request.","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fwriting-craft\u002Fskills\u002Fnarrative-structure\u002FSKILL.md",[],[],{"skillNo":78,"name":79,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":19},"01M7AT0AKGW09NKKZDYBT8VRDE","legal-brief-writing","Use when writing legal briefs, motions, or any court filing; when analyzing judicial cognition and persuasion; when applying elite brief-writing techniques from Roberts, Clement, Kagan, Garner; or when structuring legal arguments for maximum persuasion. Also use for autistic\u002Fneurodivergent advocates who need to understand how their natural communication style maps to elite legal advocacy. Triggers on any legal writing, brief drafting, motion preparation, or legal argument structuring request.","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fwriting-craft\u002Fskills\u002Flegal-brief-writing\u002FSKILL.md",[],[],{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":19},"01M7AT00VWPS4Q8B5VEMXX9NQT","journal-publishability-criteria","Use whenever a research paper, article or preprint is being written, revised, reviewed or published, and specifically before every publication of docs\u002Fpaper.md. Covers what legitimate journals actually gate on, in order of gating — the editor's desk screen (scope, novelty, package problems), what reviewers judge (methodological soundness, claims the evidence can carry, statistical and qualitative reporting), EQUATOR reporting guidelines, the ICMJE authorship criteria, COPE's view of text recycling, the near-universal generative-AI policy (no AI author, disclosure, author responsibility), TOP 2025 transparency standards, how to verify a journal is legitimate, submitting without an institution, and a pre-submission checklist. Triggers on requests to evaluate, review, prepare or submit a paper, to judge publishability, to write a declarations or data-availability section, to disclose AI use, or to check a journal.","https:\u002F\u002Fgithub.com\u002Fthe-vibey-project\u002Fvibey\u002Fblob\u002Fdevelop\u002Fsrc\u002Fvibey_tools\u002Fskills\u002Fplugins\u002Fwriting-craft\u002Fskills\u002Fjournal-publishability-criteria\u002FSKILL.md",[],[]]