Film Narrative Assistant v3
Film Narrative Assistant v3
Showcase
Description
Deeply deconstructs video scripts or course content using film narrative theory. Output includes a complete five-stage analysis: Beat identification (with original text summary), Arrangement strategy identification (using a library of 16 strategies), Narrative execution technique identification (using a library of 7 technique types), Film benchmarking, and Ultimate diagnosis. It then generates a minimalist literary criticism-style HTML visualization report (black background with serif font, single accent color, and ASCII art curves).
Why we love this skill
This skill is unique—it not only deeply analyzes film narrative structures but also generates HTML reports in a minimalist literary criticism style, perfectly combining professional analysis with elegant presentation.
Related Skills
View all
AI Short Drama Master v11.1
From a one-sentence idea to a finished video, forge a hit short drama in 10 steps (v11.0 Final Optimized · AFP 90+) 🎬 v11.0 Final Optimized Core Upgrades: • Fully restored detailed execution flow for P3-P8 • Retained v10.1 visual style selection system (6 styles throughout the entire process) • Fully restored 4 professional quick reference tables (shot size/camera movement/lighting/color) • Fully restored 6 global fallback logics + 5 P4 stage fallback logics • Added P4 quality self-check table (6 check items) • Added complete production guide for P8 final delivery package • Added anti-hallucination rules (data fidelity, source attribution, ask when info insufficient) 🔬 AFP 90+ Special Optimization (7 fixes): • Added complete definition of 'six-dimensional prompt formula' (narrative/visual/motion/emotion/technical/style) • Added 'dual-core engine': execution core generation → quality control core review, with 5 general quality controls per stage + 6 P4-specific items • Added 'environment alignment and startup check': tool availability detection, quick reference table preload, state initialization • Added 'state persistence rules': output global_state JSON block per stage, cross-round recovery without relying on memory • Added 'stage transition validation': 4 entry conditions, with modify/save/continue branch handling • Added P2/P3 negative constraints: clear prohibitions to prevent AI fabrication and deviation • Added HUD state recovery rules: automatically restore missing heartbeat signals/dashboard ✨ 10 professional stages, a complete 0-to-1 closed loop: 📋 P1 Client Five-Question Diagnosis — 3 rounds of questionnaires to lock in audience, emotion, specifications, visual style 📖 P2 Story Synopsis Generation — 200-word synopsis + character arc table + visual style adaptation + negative constraints 📚 P3 Episode Outline Splitting — 20-30 episode outlines, each with a hook + visual highlight + negative constraints 🎥 P4 Storyboard Script Generation — Six-dimensional formula + 19-column storyboard + emotion curve + 6 quality self-checks 👤 P5 Character Asset List — 6 views + expression library + action library + style keywords 🎨 P5.1 Character 6-View Generation (Optional) — AI painting generates character reference images 🏞️ P6 Scene Asset List — Cinematic lighting + color scheme + style adaptation 🖼️ P6.1 Scene Concept Art Generation (Optional) — AI painting generates scene reference images 🎞️ P7 Image-to-Video Prompts — Jimeng Seedance adaptation + style consistency 🎬 P7.1 Storyboard Video Generation (Optional) — Generate storyboard videos using image-to-video tools 📦 P8 Final Delivery Package — Complete production process guide + FAQ + publishing optimization 🔥 Deep integration of professional methodologies: • Client Five-Question Diagnosis — Precisely locate needs • Six-dimensional prompt formula — Ensure full coverage of 6 dimensions per shot • Dual-core engine — Dual guarantee with execution core + quality control core • Psychological screenwriting principles — Character arc + emotion curve • Shot composition and camera movement theory — Slow push → fast push → sudden stop golden rule • Jimeng AI delivery constraints — Dialogue ≤15 characters, frame ≤2 people ✅ 7 mandatory quality checks + 5 general quality controls + 6 P4 special self-checks ⚡ Start now and turn your short drama idea into a viral video!
Safety Short Film Diagnosis
Based on the Safety Short Film 100-Point Evaluation System V2.0 (SCI), this professional review tool provides in-depth diagnostics for film/TV scripts, short video scripts, and textual descriptions in the field of safety production. It rigorously deconstructs scripts across four core dimensions—soul, story, expression, and dissemination—helping creators precisely identify strengths and weaknesses in core themes, dramatic conflict, character development, and visual language. Through a quantitative 80-point scoring system, it generates structured diagnostic reports that include reasons for deductions and specific optimization suggestions. The tool also features a strict red-line pre-check mechanism that adeptly catches critical issues such as factual errors, violations of physical logic, value deviations, and broken emotional logic, ensuring content complies with safety technical standards and industry regulations. The review style upholds high standards and rigorous requirements, focusing not only on the script's intellectual depth and humanistic concern but also offering practical advice on dissemination dimensions such as social activation and compelling openings. This helps turn rigid safety knowledge into high-quality visual works with emotional tension and dissemination value. Whether for corporate training videos or new-media safety education, you'll find highly valuable improvement directions here.
ResearchViral Video Breakdown
Turn a popular short video into clear, learnable, and practical content methods. Whether you provide a link from Douyin, WeChat Channels, YouTube, Xiaohongshu, or another platform, a local video, screenshots, or existing dialogue and descriptions, the analysis reconstructs the material needed to examine the video itself. It also marks information that is complete or missing, so guesses are not mistaken for facts. The report includes a video profile, a time-coded transcript, structural sections, and audiovisual techniques. It further explains why the video holds viewers' attention and where it loses momentum in the middle or focus at the end. Content attribution distinguishes replicable topic and presentation methods from external factors such as account history, posting time, and platform environment. Six-dimensional scoring uses timecodes, public data, or comment evidence to evaluate hook strength, structural strength, information value, emotional tension, communication, and reach—while clearly identifying weaknesses. You will receive a Markdown breakdown report for content planning and short-video creation. It includes transferable strategies, risks that make direct imitation unsuitable, and a fill-in-the-blank script framework that can be used with replacement topics, helping you move from understanding a viral video to deciding how to shoot your next one.
Information
- Version
- v3
- Last updated
- Runtime credits
- Usage-based
- Models
- Auto
