Scroll World Landing Page
oso95/scroll-world
Builds a scroll-driven landing page where a pre-rendered camera flies through connected AI-generated scenes, using Higgsfield for stills and video clips.
Make trailers, music videos, cutscenes and page recordings for a Homie studio — a gameplay trailer captured from the studio's own running game and cut on the beat of its music; a recording of any…
$ npx skills add hashgraph-online/awesome-codex-plugins --skill video -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install hashgraph-online/awesome-codex-plugins video --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/hashgraph-online/awesome-codex-plugins.git skills-src && mkdir -p .claude/skills && cp -r skills-src/plugins/homie-rocks/homie/skills/video .claude/skills/video && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "video" agent skill from https://github.com/hashgraph-online/awesome-codex-plugins/tree/main/plugins/homie-rocks/homie/skills/video into .claude/skills/video/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "video", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/hashgraph-online/awesome-codex-plugins/tree/main/plugins/homie-rocks/homie/skills/videoType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add hashgraph-online/awesome-codex-plugins --skill video -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install hashgraph-online/awesome-codex-plugins video --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/hashgraph-online/awesome-codex-plugins.git skills-src && mkdir -p .agents/skills && cp -r skills-src/plugins/homie-rocks/homie/skills/video .agents/skills/video && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "video" agent skill from https://github.com/hashgraph-online/awesome-codex-plugins/tree/main/plugins/homie-rocks/homie/skills/video into .agents/skills/video/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "video", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add hashgraph-online/awesome-codex-plugins --skill video -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install hashgraph-online/awesome-codex-plugins video --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/hashgraph-online/awesome-codex-plugins.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/plugins/homie-rocks/homie/skills/video .cursor/skills/video && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "video" agent skill from https://github.com/hashgraph-online/awesome-codex-plugins/tree/main/plugins/homie-rocks/homie/skills/video into .cursor/skills/video/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "video", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/hashgraph-online/awesome-codex-plugins.git --path plugins/homie-rocks/homie/skills/video--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add hashgraph-online/awesome-codex-plugins --skill video -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install hashgraph-online/awesome-codex-plugins video --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/hashgraph-online/awesome-codex-plugins.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/plugins/homie-rocks/homie/skills/video .gemini/skills/video && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "video" agent skill from https://github.com/hashgraph-online/awesome-codex-plugins/tree/main/plugins/homie-rocks/homie/skills/video into .gemini/skills/video/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "video", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install hashgraph-online/awesome-codex-plugins videoInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add hashgraph-online/awesome-codex-plugins --skill video -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/hashgraph-online/awesome-codex-plugins.git skills-src && mkdir -p .github/skills && cp -r skills-src/plugins/homie-rocks/homie/skills/video .github/skills/video && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "video" agent skill from https://github.com/hashgraph-online/awesome-codex-plugins/tree/main/plugins/homie-rocks/homie/skills/video into .github/skills/video/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "video", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add hashgraph-online/awesome-codex-plugins --skill video -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install hashgraph-online/awesome-codex-plugins video --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/hashgraph-online/awesome-codex-plugins.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/plugins/homie-rocks/homie/skills/video .opencode/skills/video && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "video" agent skill from https://github.com/hashgraph-online/awesome-codex-plugins/tree/main/plugins/homie-rocks/homie/skills/video into .opencode/skills/video/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "video", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
videoMake trailers, music videos, cutscenes and page recordings for a Homie studio — a gameplay trailer captured from the studio's own running game and cut on the beat of its music; a recording of any…
Video is an agent skill from hashgraph-online/awesome-codex-plugins. Make trailers, music videos, cutscenes and page recordings for a Homie studio — a gameplay trailer captured from the studio's own running game and cut on the beat of its music; a recording of any web page while a script drives it (clicks, taps, keys, typing, scrolls, waits; Play and a few moves on the studio's own game page; a site walkthrough or product demo) in real time with honest frames; or generated footage from fal (Seedance and friends) through the creator's OWN fal account under a hard budget with a…
Its SKILL.md is about 4.5k tokens, which your agent loads only when the skill is triggered. The skill folder holds 30 other files, including scripts and reference files (for example `agents/openai.yaml`, `references/DELIVERY.md` and `references/HONESTY.md`). Compatibility notes: Node 22, ffmpeg and Chrome. Generated footage uses the creator's own fal account (FALKEY); fal's own MCP server (https://mcp.fal.ai/mcp-relay) finds models…
It sits in Media & Creative, covering AI video generation. It works with Seedance and JavaScript. The repository describes itself as: A curated list of awesome OpenAI Codex / ChatGPT plugins, skills, and resources. The 1 Codex Marketplace. See live plugins at: https://hol.org/plugins/best-codex-plugins. The licence is Apache-2.0.
9 steps, taken from the first numbered list in SKILL.md.
Read from SKILL.md and the folder at commit 78497e5. It shows what the files ask for, not the result of running them.
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Ships 6 files in scripts/ (JavaScript, from the files we listed), which the agent can run.
Shell commands in SKILL.md call:
nodenpmnpxFrom the folder's file list and the shell code blocks in SKILL.md.
Links to these hosts (documentation or services it may open):
mcp.fal.aifal.aiFrom URLs in SKILL.md, links to its own repository left out.
Names these keys or tokens, usually read from environment variables:
FAL_KEYFrom names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Node 22, ffmpeg and Chrome. Generated footage uses the creator's own fal account (FAL_KEY); fal's own MCP server (https://mcp.fal.ai/mcp-relay) finds models and reads prices.
From compatibility in the SKILL.md frontmatter.
Video loads about 4.5k tokens when it runs, and up to ~13k if it reads all its reference files. Until then it costs about 247 tokens; SKILL.md has 2,404 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.
The full file from hashgraph-online/awesome-codex-plugins at commit 78497e5, republished under its Apache-2.0 licence (© hashgraph-online). 2,404 words, ~4,497 tokens.
.claude/skills/video/SKILL.md (or your agent's skills folder). This skill also uses 25 other files; get the full folder from GitHub.A studio is the folder with studio.json. Videos live in videos/<slug>/, listed in
videos/manifest.json; a published entry is a page at /videos/<slug>/ on the studio's site,
playing the 16:9 cut (the 9:16 cut on a portrait phone). Everything runs through
scripts/video.mjs in this skill's folder (Claude Code:
node "${CLAUDE_PLUGIN_ROOT}/skills/video/scripts/video.mjs" <command>), from inside the studio.
It needs Node 22, ffmpeg and Chrome; --json on every command.
Check disk before any capture or render (df -h .): keep 10 GB free. Raw frames are deleted
once encoded. Finish in this turn: start the local site with the app's background-task tool
(not nohup ... & inside a command, which dies with the command) and wait on it.
references/HONESTY.md has them in full. In short:
references/TRAILER.md.references/METHOD.md.references/RECORD.md.Both start with node <video.mjs> check (ffmpeg, Chrome, the studio; fal only if a key is set).
One command films the game, rebuilds its sound, picks the shots and delivers three shapes:
npm run dev # background task; the site at http://127.0.0.1:8787
node <video.mjs> trailer <slug> --game <id> --url http://127.0.0.1:8787 --seconds 45 --length 20 --title "<GAME NAME>" --end "Play free"(npx --no-install homie-studio trailer <id> … is the same command, into videos/<id>-trailer/.)
requestAnimationFrame, performance.now, Date, setTimeout, setInterval and its CSS and Web
animations all move on by exactly 1/30 s, then the frame is copied. No frame is ever held, however
heavy the game or slow the computer (heldFrames is 0; speed says how it compared with real time).sound skill's player (sound.js) logs every sound it
schedules when a recorder asks; the files it names are fetched from the game's own address and mixed
again on the film's clock (work/capture/sound.wav, and sfx.wav and music.wav apart). A game
whose sound does not go through sound.js (and does not push the same events) comes out silent,
and the result says so: use capture below for those, which records the speaker in real time.shotsPlayed says what
is in each shot. work/edl.json is the edit: change it and run cut <slug> --square.<slug>.mp4 (16:9), <slug>-square.mp4 (1:1), <slug>-vertical.mp4 (9:16), a poster, an end
card (and a title card with --title) in the game's look. --keep-capture edits the film already
made (another --length, another --end) without filming again. --steps <file> presses things
while it films (references/RECORD.md; only wait, waitFor, click, tap, key, keys, type, focus).<video>/<audio> elements, Web
Workers, and the network. The default page is the big screen of a live room: a relay keeps real
time, so when frames render slower than real time (speed well under 1) the room runs fast in the
film, and the result warns. Then render smaller (--scale 0.5) or film in real time with capture.
--page <path> films another page of the site (one whose game runs in the page itself is exact).sheet), check (qa.mjs on each file), and add. Detail: references/TRAILER.md.By hand, in real time, with the studio's own song as the bed:
npm run dev # background task; the site at http://127.0.0.1:8787
node <video.mjs> capture <slug> --game <id> --url http://127.0.0.1:8787 --seconds 60
node <video.mjs> card <slug> --name title --text "<GAME NAME>" --sub "<one line>" --game <id> # the game's palette and font
node <video.mjs> card <slug> --name end --text "Play free" --sub "<site>/<id>/play" --small "Real gameplay. Empty seats are filled by bots."
node <video.mjs> edl <slug> --length 30 --song <music slug> --title "<GAME NAME>"
node <video.mjs> cut <slug>
node <video.mjs> sync <slug>
node <video.mjs> sheet <slug> --in videos/<slug>/<slug>.mp4npx --no-install homie-studio dev --stop
(it stops exactly this studio's server), never with a name pattern like pkill -f wrangler.npx --no-install homie-studio dev --port <free port> and use that address.capture records the game's big screen (/<id>/tv: a spectator in a live public room, bots in
empty seats) off a headless GPU Chrome for real seconds, with the game's own WebAudio sound on
the same clock (the browser is muted; nothing plays out loud). It presses nothing. The live
site works too (--url the studio's address); a local run keeps strangers out of the shot.
A heavy game paints few frames at 1920x1080 (it prints "fps from the page" and a note when that
is well under the film's 30): capture again with --scale 0.67 (a 1280x720 page, scaled up to
1920x1080 when it is encoded) or a lower --fps. One game measured 8 fps at 1080p and 39 at 720p.--song, from the music skill) or the game's sound alone.
No song yet: the sound skill synthesizes a theme for free (its bar grid cuts the trailer the same
way), or the music skill renders one with ElevenLabs (credits: ask).edl finds the song's bar lines, picks the busiest moments of the capture (motion, not guesses),
and writes work/edl.json: a title card of one bar, shots of one bar each, the end card on what
is left. Title and end cards made with card … --game <id> wear the game's palette and display font
(games/<id>/style.json, the style skill's decisions). Every cut lands on a bar line. Read it and change it: order, which moments, where the
song starts (--bed-from-bar).cut renders both deliveries: 16:9 (1920x1080) and 9:16 (1080x1920, the whole game frame over a
blurred fill, never a crop that hides the action), and 1:1 (1080x1080) with --square; 30 fps, the
song and the game's sound mixed, -14 LUFS, faststart, and a poster frame. Every cut is on the frame
nearest its beat (segments are counted in frames, so cuts never drift from the music: cuts lists
each one's frame), and the picture is yuv420p, limited range, converted and tagged BT.709 (a
capture's frames are full range; cut fails if the file does not read back that way).sync checks every cut: the picture must change within one frame of the sound's onset.
sheet makes a contact sheet: open it and look before anyone else sees the video.node <skill folder>/scripts/qa.mjs videos/<slug>/<slug>.mp4 checks the delivered file itself: every
frame decodes, phone-safe encoding, faststart, no black or frozen stretches, loudness and true peak,
under the site's 25 MiB. Fix every FAIL before add. Delivery, capture and honesty traps:
references/DELIVERY.md.The provider, only now. fal through the person's own account: they create a key at
https://fal.ai/dashboard/keys and set FAL_KEY in the environment your AI runs in
(never pasted into the chat, never written into the studio). check tests it for free.
fal's own MCP server is the way to find models and read their input schemas and prices
(search_models, recommend_model, get_model_schema, get_pricing): use it rather than memory,
because video endpoints retire and reprice monthly. Not connected? Offer it, and the person signs in
on fal's own page (no key): in Claude Code claude mcp add --transport http --scope user fal https://mcp.fal.ai/mcp-relay, then /mcp; in Codex codex mcp add fal --url https://mcp.fal.ai/mcp-relay, then codex mcp login fal. Paid calls go through video.mjs gen,
never the MCP's run_model or submit_job: gen prices, caps, receipts and resumes; the MCP call
does none of that (the Homie mod in Claude Code, and Homie's hooks in Codex, hold one as a call whose cost cannot be read first).
The budget, before the first call. Plan the shots, price one of each kind, add them up, and ask the person for a cap in dollars:
node <video.mjs> price --model <fal model> --input <input.json> # free: fal's own unit price x this input
node <video.mjs> budget <slug> --cap <dollars they agreed to>
node <video.mjs> gen <slug> --model <model> --input <input.json> --out work/<file> --dry-run
node <video.mjs> gen <slug> --model <model> --input <input.json> --out work/<file> --yesgen refuses with no budget, refuses past the cap, and asks again without --yes. The receipt is
written the moment fal accepts the job (videos/<slug>/budget.json, videos/receipts.jsonl).
Running the same gen again resumes the job from <out>.request; it never pays twice. Strings
"@file:<path>" in the input are uploaded to fal's storage first (free). A model whose billing
unit the script cannot count is refused, not guessed: price it by hand from the model page. Spend
the riskiest shot first (the pilot), look at it, then the rest.
The method (references/METHOD.md has every step):
videos/<slug>/STYLE.md): palette, line, type, what the look risks.node <video.mjs> grid <slug> --song <music slug> gives the
song's beats, bars and every sung word's time; each shot starts and ends on a bar line.slice), so mouths and hits land on the music. 480p or 720p is
enough under a draw-over.lag <slug> --base <clip> --ref <cut>. A clip that sings its
reference plays it back at 0 ms; anything else is regenerated or shifted.film init <slug>, film frames <slug> --base <clip> --id <shot>,
edit film/shots.json and film/look.js, then film render <slug> (and --mode v for 9:16).
Every frame is a function of time, rendered frame-exactly in a headless browser.words <slug> --in <cut> (a frame at every sung word, labelled) and
sync <slug> --in <cut> --at <hit times>. Fix and re-render until every word and hit lands.sheet <slug> --in <cut> --every 1. Look at every cell. Then
scripts/qa.mjs <cut> on each delivery (references/DELIVERY.md: continuity, faces, green screens).npm run dev # background task; the site at http://127.0.0.1:8787
node <video.mjs> record <slug> --steps <steps.json> --url http://127.0.0.1:8787 [--device phone] [--name <take>]
node <video.mjs> sheet <slug> --in videos/<slug>/work/record/recording.mp4click, tap, drag,
hover, key, keys, type, scroll, goto, wait, and waitFor a selector, a text or a state
("js"), in the page or in the game's own frame ("frame": "game"). Any step can carry a
"caption". references/examples/studio-play.json records a studio's own game from its landing:
Play, then the arrow keys; studio-play-phone.json the same on a phone with the game's touch stick.
Both are tested against a real studio. Every step: references/RECORD.md. A Game Lab's side by side
(New beside Today, slowed down, stepped through its phases) is a page too:
the lab skill's references/record-lab.json records it from the lab's own address.--no-cursor drops them); the take's honesty line says so.recording.json has each step's second, what its frame received (keys, presses, touches: delivered,
not judged), the page's and the game's real frame rates and the renderer. Under 20 fps or on a
software renderer it warns: that take shows a slow computer, not the page.sheet and look; add <slug> --file videos/<slug>/work/record/recording.mp4 --kind clip --title "…" (copy its captions.vtt to videos/<slug>/captions.vtt first) and publish. Or cut
it into a trailer like any capture.node <skill folder>/scripts/record-page.mjs --steps <steps.json> --out <folder>
records any address (it needs Chrome, ffmpeg and puppeteer-core).A published trailer with --for-game <id> plays, muted, in the hero of that game's landing page when the
game has no hero loop of its own, and the landing links to it ("Watch the trailer"). The best hero is a
short silent loop of real play in games/<id>/hero/ (the game skill, "Its landing page", cuts it from a
capture).
node <video.mjs> add <slug> --title "<Title>" --blurb "<one line>" --kind trailer|music-video|cutscene [--for-game <id>] [--for-song <music slug>] --publish
node <video.mjs> publish <slug>A studio needs no storage for this: the site serves each file itself, up to 25 MiB a file, and
cut and film render hold their deliveries under that (a longer video gets a lower bitrate).
Bigger masters stay in the studio folder. A studio with storage (storage add: R2 on the studio's
own Cloudflare account, which asks the person for a payment method first) keeps its big media there
by default: the deploy uploads each delivery (over 1 MiB, or left out of git), reads it back and
checks it by SHA-256 before the site stops carrying it, and the page plays it at the same address.
Cuts can then run at 12 Mbit/s. R2 has no egress fees; storage is free up to 10 GB-month, then
US$0.015 per GB-month.
add writes the manifest entry: the 16:9 cut the player plays, the 9:16 cut, the poster,
captions if there is a captions.vtt, the credits (the models used, the song's credits), the
honesty line from the capture (or the recording), and the money spent. publish redeploys (with
storage, the deploy moves the big files to R2 first, each checked by SHA-256), and checks the page
(200) and the video (a byte range, 206: phones seek with them) at its address. A studio not online
yet: the publish skill first.
The page link, the length, both files, what it cost (from the receipts) against the cap, what
is generated and what is captured, and what the sync check and the contact sheet showed. Commit
videos/ (big media stays out of git; the manifest, edl, shot list, film code and receipts go in).
© hashgraph-online, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
SKILL.md and 25 other files (scripts, references) in plugins/homie-rocks/homie/skills/video of hashgraph-online/awesome-codex-plugins.
Open the folder on GitHubat commit 78497e5
Video next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Video this skillhashgraph-online/awesome-codex-plugins | 1.2k | — | ~4.5k | Automated safety check: Pass | Apache-2.0 | |
| Scroll World Landing Pageoso95/scroll-world | 9.8k | 1 repos | ~12k | Automated safety check: Notes | MIT | |
| Seedance Storyboard Generatorliangdabiao/Seedance2-Storyboard-Generator | 2.5k | — | ~2.2k | Automated safety check: Pass | None | |
| Libtv Videonexu-io/nexu | 3.3k | — | ~3.5k | Automated safety check: Pass | MIT | |
| Analyze Videokrusemediallc/arcads-claude-code | 1.6k | — | ~4.2k | Automated safety check: Pass | MIT | |
| Imagesmixs/visual-skills | 484 | — | ~2.1k | Automated safety check: Pass | CC-BY-4.0 |
oso95/scroll-world
Builds a scroll-driven landing page where a pre-rendered camera flies through connected AI-generated scenes, using Higgsfield for stills and video clips.
liangdabiao/Seedance2-Storyboard-Generator
专业的Seedance 2.0平台AI视频脚本和分镜生成器。当用户要求:(1) 将文章/故事转换为视频脚本,(2) 生成Seedance 2.0分镜提示词,(3) 规划多集AI视频系列,(4) 为GPT-Image-2、Seedream、Nano Banana…
nexu-io/nexu
Seedance 2.0 video & image generation via LibTV Gateway - AI text-to-video, image-to-video, video continuation, style transfer, and text-to-image using Seedance 2.0 model.
krusemediallc/arcads-claude-code
Analyze a reference video and reverse-engineer its style into a reusable Seedance 2.0 prompting template.
smixs/visual-skills
Image prompting skill for Nano Banana (NBP/NB2) and GPT Image 2.5 (Flare/Sunburst).
cclank/lanshu-waytovideo
使用剪映(Jianying/小云雀)的 Seedance 2.0 模型自动生成AI视频。支持文生视频(T2V)、图生视频(I2V)、参考视频生成(V2V)和向后延伸(Extend)四种模式。当用户需要生成AI视频、使用Seedance模型创作短片、基于参考图像/视频进行风格转换,或对已有结果继续延长时使用此技能。需要预先配置 cookies.json 登录凭证。
hashgraph-online/awesome-codex-plugins
Create original anime-style reaction stickers as looping GIFs and MP4 previews, using generated character pose sheets and timed key poses.
hashgraph-online/awesome-codex-plugins
Manage and query Calibre libraries with the calibredb CLI (local paths or Calibre Content server URLs).
hashgraph-online/awesome-codex-plugins
A skill your agent uses when adding, changing, testing, or debugging Rust HTTP APIs and services, especially when Codex needs black-box integration tests, random-port app startup, real database test…
hashgraph-online/awesome-codex-plugins
Make a studio's game look like something at build time — a cover from a real frame of the game (free), painted covers, backdrops, textures and character plates from image models through the…
hashgraph-online/awesome-codex-plugins
Balance game difficulty, resources, rewards, probability, progression, economies, and dominant strategies.
hashgraph-online/awesome-codex-plugins
Analyze nonfiction manuscripts for reader engagement signals, including heading-level word counts, slow starts, long slogs, weak takeaway titles, value pacing, beta-reader comment dropoff, and…
Works with
Categories
Make trailers, music videos, cutscenes and page recordings for a Homie studio — a gameplay trailer captured from the studio's own running game and cut on the beat of its music; a recording of any…. Video is an agent skill from hashgraph-online/awesome-codex-plugins.
Video fits situations like: someone in a studio asks for a trailer; A clip for social media; A video of my game; tutorial video of a site.
Run `npx skills add hashgraph-online/awesome-codex-plugins --skill video -a claude-code`. Or copy the skill folder (plugins/homie-rocks/homie/skills/video in hashgraph-online/awesome-codex-plugins) into .claude/skills/video in your project. Claude Code loads it when a task matches its description.
Run `npx skills add hashgraph-online/awesome-codex-plugins --skill video -a codex`. Or copy the skill folder (plugins/homie-rocks/homie/skills/video in hashgraph-online/awesome-codex-plugins) into .agents/skills/video in your project. Codex loads it when a task matches its description.
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add hashgraph-online/awesome-codex-plugins --skill video -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/video, .gemini/skills/video, .github/skills/video and .opencode/skills/video in your project.
Going by SKILL.md and its folder, Video needs JavaScript for the scripts in its folder, the command-line tools its instructions call (node, npm and npx) and credentials named FAL_KEY. Our summary lists: Node.js; A credential in FAL_KEY. Compatibility (from SKILL.md): Node 22, ffmpeg and Chrome. Generated footage uses the creator's own fal account (FAL_KEY); fal's own MCP server (https://mcp.fal.ai/mcp-relay) finds models and reads prices..
SKILL.md names 2 domains. As links in the text: mcp.fal.ai and fal.ai. This is read from the text; nothing was executed.
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. The check reads SKILL.md only: the scripts in the folder are not scanned, so read them before running anything.
Video is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
About 4.5k tokens (SKILL.md is roughly 18k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 8.5k tokens, read only when the agent opens those files.
Skills that share tags, products or a category with Video: Scroll World Landing Page (oso95/scroll-world, 9.8k stars), Seedance Storyboard Generator (liangdabiao/Seedance2-Storyboard-Generator, 2.5k stars), Libtv Video (nexu-io/nexu, 3.3k stars) and Analyze Video (krusemediallc/arcads-claude-code, 1.6k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
hashgraph-online (a GitHub organization) maintains it in hashgraph-online/awesome-codex-plugins, which has 1,242 GitHub stars. The repository holds 686 skills in this directory. The repository was last updated on October 8, 2026.
Source: hashgraph-online/awesome-codex-plugins on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.