Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add phileiny/h3-storyboard-skill --skill h3-storyboardgit clone --depth 1 https://github.com/phileiny/h3-storyboard-skillWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/phileiny/h3-storyboard-skill/h3-storyboard)<a href="https://agentmods.dev/skills/phileiny/h3-storyboard-skill/h3-storyboard"><img src="https://agentmods.dev/badge/skills/phileiny/h3-storyboard-skill/h3-storyboard/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/phileiny/h3-storyboard-skill/h3-storyboard"><img src="https://agentmods.dev/badge/skills/phileiny/h3-storyboard-skill/h3-storyboard.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00114 | $0.10080 |
| Opus 5 | $0.00057 | $0.05040 |
| Sonnet 5 | $0.00023 | $0.02016 |
| Haiku 4.5 | $0.00011 | $0.01008 |
Grade A, and why
h3-storyboard scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 753 lines — stays where its author put it; the contents beside it link to each section on GitHub.
H3 分鏡與表演
minimax-h3 那份 skill 教你怎麼把已經想好的東西寫成 H3 格式。
這份教你怎麼想——劇本怎麼拆成鏡頭、情緒怎麼變成模型演得出來的動作。
官方和社群的 skill 都停在翻譯層,沒有人做這段(instann/minimax-h3-director
把「自動拆鏡」列在 roadmap,還沒做)。下面全部是實拍歸納出來的,
每一條都標了驗證狀態。
一、🔴 第一定律:一顆鏡頭裝不下太多節拍
在一顆鏡頭裡塞太多表情節拍,模型會往「平均運動」收斂,把它們全部抹平。 寫了睜眼、揚眉、咬下唇,畫面上什麼都沒發生——而且不會有任何報錯。
三個版本的對照實驗,同一顆震驚戲、同一顆 seed、同樣的臉部節拍:
| 鏡頭結構 | 對白 | 情緒峰值的 PSNR | 結果 | |
|---|---|---|---|---|
| A | 一顆 7 秒特寫,9 個節拍 | 無 | 37–42 dB | 臉完全沒動 |
| B | 拆成三顆 2–3 秒,各一個主節拍 | 無 | 22–23 dB | 表情全部到位 |
| C | 同 B | 加 <d> |
19 dB | 表演幅度再大一些 |
(PSNR 越低表示畫面變化越大。42 dB 等於凍結幀。)
A → B 是主因:拆鏡頭。 從 40 掉到 23,絕大部分的效果在這一步。 B → C 是加成:對白。 從 23 再到 19,有幫助但不是機制。
所以怎麼做
① 先數節拍。 一顆鏡頭超過兩三個表情節拍就要拆。
② 拆成 2–3 秒的短鏡頭,每顆一個主節拍。
③ 有台詞就寫進去(見下),沒有也能成立。
切鏡本身就是表演——觀眾在切點會自動重新讀取角色狀態, 所以拆開不只是為了讓模型執行得到,也是為了讓觀眾看得到。
驗證:2026-08-26,Ref2VA 243 幀,三版對照,seed 固定。 這是隔離過的:A→B 只改鏡頭結構,B→C 只改一句台詞。
旁證:B 站有支純本地實測是「統一首幀 · 10 種情緒 · 一支 6 秒一種情緒」, 同一個結論——一支只裝一種情緒。
<d> 的角色:它偷時間
有台詞的鏡頭表演幅度更大,但機制不是「對白讓臉會動」—— 是 H3 把整支影片的時間重新分配,多給了有台詞的那一鏡。
同一支 243 幀的四鏡影片,只差一句 VO 對白(其餘完全相同:同 seed、
同參考圖、同 ref_image_size=max):
Shot 1 Shot 2 Shot 3(有台詞) Shot 4
提示詞規格 77 幀 48 幀 74 幀 44 幀
無台詞 89 43 55 ❌ 56
有台詞 88 29 84 42
無台詞版把情緒峰值那鏡壓縮到 55 幀,比規格少了 19 幀,演到一半就切走。 有台詞版給了 84 幀,臉部最大變化 28.5 dB(無台詞 30.2 dB)。
代價在別的鏡頭身上。 Shot 4 被壓到 42 幀之後:
無台詞:配角在落地窗外還在走動,陽台雨遮、欄杆都保留
有台詞:配角完全不存在,落地窗退化成一扇普通窗戶
Shot 1 建立的場景元素,有台詞版沒有延續到 Shot 4。
所以:
- 要表演的鏡頭 → 寫台詞(內心獨白也算,寫成旁白版)
- 要背景延續性的鏡頭 → 不要跟台詞放在同一支影片裡,或者兩版都跑再剪
內心獨白的寫法:
the young woman, in a small unsteady voice that catches once partway through (S1),
says in an off-screen voiceover: <d>[Chinese]……</d>
while her lips remain completely closed.
情緒寫在 delivery 欄位,不是寫在 <d> 裡面。
官方規格:識別語、ID、動作、delivery 都在 <d> 外面,<d> 內只放語言標籤和逐字台詞。
但不要為了表演硬塞台詞。 設計上就不說話的角色(在聽、在迴避、在生悶氣), 拆短鏡頭就夠了——B 組證明沒有台詞一樣演得出來。
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 12d ago First seen · 753 lines · 114 tokens per session scan A 21580ba97556
h3-storyboard is a skill published in the GitHub repository phileiny/h3-storyboard-skill (161 stars, last pushed 15d ago), licensed MIT. It adds 114 tokens to every session and 10,080 once invoked, about $0.0006 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…