video_toolkit
Make short explainer videos from a brief with claude-code-video-toolkit — submit a scene spec, poll a job, get a QC-checked MP4. Also covers the individual tools (AI voiceover, images, clips, music, talking heads) for custom work.
Install / Use
npx skills add digitalsamba/claude-code-video-toolkit --skill video-toolkitInstalls into whichever agent you are using.
SKILL.md
Installable skill definition
Quality Score
Category
MarketingSupported Platforms
Our assessment of video_toolkit
video_toolkit scores 94/100 on our quality scale, 96th of 646 Marketing skills we index (top 15%).
Its SKILL.md is 36 KB long, well organised into 31 sections with 14 code examples: a thorough specification that gives an agent plenty to work with.
With 2,544 GitHub stars, it is one of the more widely adopted skills in the catalogue.
Maintenance, license and trust
- The repository was last updated 5 days ago, so video_toolkit is actively maintained.
- It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
- Its trust signals score 100/100, with no cautions. These come from repository metadata, not a code audit — read the skill file before letting an agent act on it.
video_toolkit compared with similar skills
All 4 of these similar skills score higher than video_toolkit; compare them before choosing.
| Skill | Score | Stars | Updated | Format |
|---|---|---|---|---|
| video_toolkit (this skill)by digitalsamba | 94 | 2.5k | 5d ago | SKILL.md |
| taste-skillby Leonxlnx | 100 | 117.6k | 9d ago | CLAUDE.md |
| Agent-Reachby Panniantong | 100 | 112.7k | 9d ago | CLAUDE.md |
| LocalAIby mudler | 100 | 50.4k | today | MCP Server |
| x402by browser-use | 100 | 121.0k | 4d ago | SKILL.md |
Frequently asked questions
- How do I install video_toolkit?
- Run
npx skills add digitalsamba/claude-code-video-toolkit --skill video_toolkit. The install tabs above show the steps for each supported agent. - Which AI agents does video_toolkit work with?
- It is written for Claude Code, as a SKILL.md file. Other agents that read the same format can often use it too.
- Is video_toolkit safe to use?
- It is MIT-licensed and scores 100/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
- Is video_toolkit still maintained?
- The repository was last updated 5 days ago, so video_toolkit is actively maintained.
Skill content
View source on GitHubname: video_toolkit description: Make short explainer videos from a brief with claude-code-video-toolkit — submit a scene spec, poll a job, get a QC-checked MP4. Also covers the individual tools (AI voiceover, images, clips, music, talking heads) for custom work. metadata: openclaw: emoji: "🎬" skillKey: "video-toolkit" os: ["darwin", "linux", "win32"] requires: bins: ["uv", "ffmpeg"]
Video Toolkit
Turn a brief into a finished MP4 with claude-code-video-toolkit.
This skill is written for any agent runtime: Claude Code, OpenClaw, Codex, Hermes, a bare API loop with a
run_tool call, or an MCP client. Nothing here depends on how your runtime backgrounds work, because the long
work runs in a detached job you poll.
Who this is for
You are an agent: someone asks you for a video through a channel (a chat message, a Discord or Slack thread, an MCP client) and you make it without a person at your terminal. You report back in that channel. The person asking sees your messages and the finished video, never the toolkit.
What the toolkit can make: pick a route first
Every route has one of two support levels. catalog gives it per template as agent_support.
job: one prompt.video_job.pyruns it unattended from one spec. You write the scenes; it voices, illustrates, captions, builds and QCs them; you post the MP4. Any runtime can do this, through the CLI or MCP.session: a working session. Copy a template or example intoprojects/<name>/, edit its code with the skill it names, render with Node, look at the result, fix it, render again. That needs a Claude Code (or Codex-style) session in the toolkit, which is what the toolkit is built around. The session can be a person's, or yours, if your runtime can run one headless with a shell and Node. Either way, the skill's review points (script, a contact sheet, the render) go back to the person who asked, in their channel, before you spend more. Tell them up front what the route needs from them (for a synthetic UI video: the app's source, the click path, the narration).
A route moves from session to job once its inputs fit in a message, a spec drives the build with no code
written, QC can judge the result, and there's a cost cap. Offer a job when one would do.
| The user asks for | Route | Support | Read first | See it |
|---|---|---|---|---|
| A short explainer on a topic (Shorts, Reels, a 16:9 explainer) from a brief | concept-explainer-short via video_job.py | job | this file | examples/sky-blue-short |
| A post or short video for someone who has a kit (a small business, a coach, a product team) | video_job.py with "kit": their photos as file scenes, text scenes in their colours | job | kits/README.md | |
| A walkthrough, feature demo or release-notes video of an app whose source you can read | synthetic UI: templates/ds-release-notes, components Stage, Cursor, cues (Terminal for a shell, TerminalSession for a Claude Code, Codex or other agent CLI session) | session | .claude/skills/synthetic-ui-demo/SKILL.md | examples/ds-release-649, examples/ds-free-schedule-demo |
| A how-to or tutorial about a tool or CLI, in chapters, with Shorts cut from it | templates/howto (mark-and-track opening and close, ledger, figure, bars, split and TerminalSession scenes on narration cues, facts.json, chapters from the timeline) | session | templates/howto/README.md, .claude/skills/video-design/SKILL.md | templates/howto itself (Getting started 1) |
| The same, but you can't read the source | Playwright recording into product-demo | session | .claude/skills/playwright-recording/ | |
| A marketing or product launch video | templates/product-demo | session | the template's README | examples/digital-samba-skill-demo |
| A sprint or monthly review | templates/sprint-review, templates/sprint-review-v2 | session | the template's README | examples/hello-world |
| A data story: a number or a trend drawn as the narration talks | DataChart (LineChart, BarChart, Counter, SplitBar) in Remotion; Matplotlib + moviepy without Node | session | lib/components/DataChart.tsx, .claude/skills/moviepy/ | examples/data-chart-stars, examples/data-viz-chart |
| A fun or holiday video about the user's own group (a release night as an action-show intro, a Christmas ad) starring their photos, or a spoof of a shared topic with an invented cast | Remotion shots + LivingPhoto, cast from image_edit then ltx2.py --input (ltx_extend.py for long shots), in-jokes from the group's own chat and docs | session | .claude/skills/parody-short/SKILL.md | |
| A 3D character acting out a gag as a film of its own, or a 3D reveal or character inside another video (the scene turns out to be on a laptop; a character reacts beside the content) | Blender shot spec: tools/blender_render.py (check, then render, or render --clips --alpha for BlenderClip in Remotion). The cast is described in JSON (build, face, hair, clothes, a seed) and can be kept in a kit's cast/ folder. Needs Blender on the host: blender_render.py doctor. Experimental: a character fitted from a generated 3D model (blender_render.py fit), on its feet in a kitchen or a lecture hall, walking, sitting, carrying a mug. For a character of a real person, tools/reference_capture.py takes the reference photos at a webcam (the person runs it in a browser; it is not a job step) | session | .claude/skills/blender/, docs/blender.md | examples/blender-disk-panic, examples/blender-in-remotion, examples/blender-kitchen |
| An app, a UI or a product shown as a 3D object: a screenshot or a render on a phone or a floating card, or a 3D model (glTF, OBJ, FBX, STL) on a studio stage, with the camera orbiting or pulling back from the screen | Blender shot spec with set kind studio and product, orbit, reveal shots; render is a film of its own, --clips a clip for any build, --clips --alpha a transparent clip for a Remotion scene | session | .claude/skills/blender/, docs/blender.md | examples/blender-product-shot |
| An opening and closing ident from a logo: the SVG built in 3D, its parts moved on beats (a bar that snaps, a shape that grows, a key that is pressed), a sound on the hit, in 16:9 and 9:16; or a check that a logo holds up at small sizes | Blender shot spec with a logo prop and an ident shot (render --clips --with-sound, --alpha for a transparent clip); check compares the last frame with the flat SVG. tools/mark_sheet.py --svg mark.svg --brand NAME needs no Blender: the mark at 512 to 16 px, on each ground, in a circle, as a watermark, with warnings | session | .claude/skills/blender/, .claude/skills/video-design/idents.md, docs/blender.md | examples/blender-ident |
| A sub-30 s ad spot or a brand stinger | moviepy, audio-anchored | session | .claude/skills/moviepy/ | examples/quick-spot, examples/ds-crt-stinger |
| A presenter talking to camera | soulx.py (below), placed in a template with NarratorPiP | tool, or session | Talking heads, below | |
| Change an existing video: new voice, music, vertical cut, highlight reel, subtitles, watermark | redub.py, addmusic.py, cutdown.py, highlights.py, captions.py --srt, dewatermark.py | tool | QC, captions, cut-downs, below | |
Blender is optional, so check before offering it. uv run tools/blender_render.py doctor --no-bench exits 0
when the host has it. If it does, suggest a 3D shot wherever one fits while planning any video, not only when the
user says "3D": its clips are ordinary video files, so one can sit in a Remotion scene, in a moviepy build, or as a
file scene in a video_job.py spec. If it doesn't, leave the Blender routes out of what you offer (mark_sheet.py still works).
Say how long it takes before starting: a few seconds of the clay cast is minutes, but a film with a fitted
character in a room is about 2.2 s a frame with a GPU, two to three hours a session, over a few sessions with the
person's notes between them (docs/blender.md, "A film over several sessions").
ds-release-notes is built for one product's screens, but the method isn't: keep its Stage and Cursor
wiring and rebuild your own app's screens the way .claude/skills/synthetic-ui-demo/ describes.
The catalogue lives on disk. uv run tools/video_job.py catalog (MCP: video_catalog) lists every template
with its demo videos and the skill that drives it, the example builds, components, brands, design sources and voice
presets, read from _internal/toolkit-registry.json. When a user asks "what can you make?", answer from it, with
the demo links, rather than from memory: the list grows each release. Skills under .claude/skills/ are written
for Claude Code, but they are plain markdown: open SKILL.md in the folder and follow it.
When asked "what can you make?", answer with the job routes first (what you can do from this message), then
the session routes and what each needs from them, each with its demo link from catalog.
How it looks: brand, design source, transitions
- Brand (
brands/<name>/brand.jsoncolours and fonts,voice.jsona voice): pass"brand": "<name>"in a job spec, or--brandto the image tools.cataloglists the brands on this host. - Design direction for a session:
.claude/skills/video-design/turns a direction into scenes (colour by act, type scale, one data accent) and critiques the render. The direction comes from a design source:ours(the default, video-design's own palettes, nothing to install),taste-skill(recommended, installed separately) orfrontend-design. A project records its choice indesign.md. After a render,contact_sheet.pytiles one frame per scene for the critique. - Turning a scene in place (the same thing shown again in a new way):
PageTurns. Between scenes, a hardwipe(); the effect transitions inlib/transitions/are for brands that call for them. - Other components (
lib/components/):AnimatedBackground,SlideTransition,Label,Vignette,LogoWatermark,SplitScreen,Envelope,PointingHand,MazeDecoration,FilmGrain,Captions. Remotion itself is documented in.claude/skills/remotion-official/(synced from upstream) and the toolkit's patterns in.claude/skills/remotion/.
The rest of this file is the job route, then the individual tools.
Where the toolkit is
Every command below runs from the toolkit root. Find it, in this order:
$VIDEO_TOOLKIT_ROOTif set (also honoured by the toolkit's own scripts);$VIDEO_TOOLKIT_DIR(an alias some runtimes set);- a
claude-code-video-toolkit/directory under your workspace or home.
cd "${VIDEO_TOOLKIT_ROOT:-${VIDEO_TOOLKIT_DIR:-$HOME/claude-code-video-toolkit}}"
uv run tools/verify_setup.py --json # prerequisites, keys and endpoints; [x] means ready
Never run a tool from inside a project directory: tools resolve paths relative to the toolkit root. If your
runtime can't cd, pass --directory: uv run --directory /path/to/toolkit tools/video_job.py ....
No shell at all? Use the MCP server instead: uv run --directory /path/to/toolkit --extra mcp tools/video_mcp.py
exposes video_estimate, video_submit, video_status, video_result and video_list_presets over stdio.
Same flow as below, same JSON.
The quick path: estimate → submit → poll → result
tools/video_job.py is one async entrypoint. submit returns in seconds with a job id; a detached worker runs
the whole pipeline (voiceover → visuals → captions → build → QC); every later call is short and reads state
from disk. That is how it fits a per-call time limit, and it survives you going away between turns.
1. Write the spec (you are the LLM)
A spec is a list of scenes, each with narrat
Truncated for display — read the full file on GitHub.
Related Skills
taste-skill
117.6kAgent skills that give coding agents taste: art direction, reference images and a browser check instead of generic AI pages
Agent-Reach
112.7kGive your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.
LocalAI
50.4kLocalAI is the open-source AI engine. Run any model - LLMs, vision, voice, image, video - on any hardware. No GPU required.
x402
121.0kSet up Browser Use Cloud payments with x402 — pay per request from a crypto wallet (USDC on Base mainnet), no signup or API key.
Languages
Trust signals
From repository metadata: license, adoption, age and documentation. Not a code audit — see the Safety scan above for what the skill file itself contains.
