Agent skill

Cohub Generate

by netaart in netaart/cohub

Generate or transform images, video, speech, and music with Cohub multimodal models via cohub generate.

Apache-2.0Auto-check passedMedia & Creative

Install Cohub Generate

skills CLI
$ npx skills add netaart/cohub --skill cohub-generate -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install netaart/cohub cohub-generate --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/netaart/cohub.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/cohub-generate .claude/skills/cohub-generate && rm -rf skills-src

Use ~/.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/

Facts

Skill name
cohub-generate
GitHub stars
570
Token cost
~1.2k tokens
SKILL.md length
328 words
Files
1
Skills in repo
29
Repo updated
First seen
Licence
Apache-2.0

At a glance

Generate or transform images, video, speech, and music with Cohub multimodal models via cohub generate.

  • The user asks to create
  • SKILL.md covers Installation, Models, Generate and Speech (TTS), plus 1 more section
  • Calls npm
  • Remove backgrounds

What it does

Cohub Generate is an agent skill from netaart/cohub. Generate or transform images, video, speech, and music with Cohub multimodal models via cohub generate. Use when the user asks to create, edit, restyle, animate, remove backgrounds, synthesize speech (TTS), or generate songs.

Its SKILL.md is about 1.2k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Media & Creative, covering Text to speech and voice and Image editing. The repository describes itself as: A living space where people and agents create, play, and build together. The licence is Apache-2.0.

When your agent uses it

  • The user asks to create
  • Remove backgrounds
  • Synthesize speech (TTS)

Example prompts

  • “/cohub-generate”

Requirements

  • Node.js

What it can do on your machine

Read from SKILL.md and the folder at commit 1a4b7c1. It shows what the files ask for, not the result of running them.

  • Tool permissions

    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.

  • Runs code

    Shell commands in SKILL.md call:

    • npm

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use npm, which can reach the network depending on how they are called.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Cohub Generate loads about 1.2k tokens when it runs. Until then it costs about 61 tokens; SKILL.md has 328 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~61
When it runs · the whole SKILL.md, loaded when a task matches
~1.2k

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.

Safety

Auto-check passed

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); files beside SKILL.md are not scanned.

SKILL.md

The full file from netaart/cohub at commit 1a4b7c1, republished under its Apache-2.0 licence (© netaart). 328 words, ~1,176 tokens.

Download SKILL.mdSave it as .claude/skills/cohub-generate/SKILL.md (or your agent's skills folder).
name
cohub-generate
description
Generate or transform images, video, speech, and music with Cohub multimodal models via `cohub generate`. Use when the user asks to create, edit, restyle, animate, remove backgrounds, synthesize speech (TTS), or generate songs.

Cohub Multimodal Generation

Use cohub generate to create or transform images, video, speech, and music.

Prefer simple, explicit commands. Use --json when reading output for decisions, extracting URLs, or chaining commands. If a target Space is required, add -s "$COHUB_SPACE_ID" or an explicit Space ID.

Installation

If cohub is unavailable, install it and check availability:

bash
npm install -g @neta-art/cohub-cli
cohub --help

Models

Always resolve models from the live list; do not hardcode a catalog.

bash
cohub models ls --model-type multimodal --json

Typical categories: text-to-image / image editing, text-to-video / image-to-video, text-to-speech (TTS), background removal, music.

Before using non-default parameters or reference media, inspect the model schema for supported inputs, roles, parameters, defaults, and examples:

bash
cohub models show <model> --json

Generate

Default to returning generated result links directly, with Markdown preview when possible (![alt](url) for images, direct links for video and audio). Save locally only when the user asks for a file download, local editing, or post-processing.

Generate from text:

bash
cohub generate "a calm lake at sunrise" \
  --model <model>

Generate with reference media:

bash
cohub generate "restyle this image" \
  --model <model> \
  --image ./input.png \
  --param size=1024x1024

Supported inputs: --image, --video, and --audio, each repeatable. Pass a URL or a local path; local files upload to an unlisted public URL first, so tasks store a reference instead of inline data. For files that must stay private, add --inline to keep them inside the task.

When a model requires input roles, prefix the path or URL:

bash
cohub generate "smooth transition between two shots" \
  --model <model> \
  --image first_frame=https://example.com/first.png \
  --image last_frame=https://example.com/last.png

cohub generate "keep these characters consistent" \
  --model <model> \
  --image reference_image=https://example.com/a.png \
  --image reference_image=https://example.com/b.png

cohub generate "lip-sync to this spoken take" \
  --model <model> \
  --image reference_image=https://example.com/portrait.png \
  --audio reference_audio=https://example.com/speech.mp3

Roles include first_frame, last_frame, reference_image, reference_video, and reference_audio. Check models show for what a model accepts. Do not mix first/last frame roles with reference roles. Seedance 2 reference audio needs an image or video in the same request.

Pass generation parameters with --param key=value (repeatable; JSON, number, or boolean values) or --parameters '<json>':

bash
cohub generate "cinematic drone shot over misty mountains" \
  --model <model> \
  --param duration=5 \
  --param resolution=720p \
  --param ratio=16:9

Results print their media facts when available, and videos their last frame:

text
video 720×1280 · 11.0s: https://…/clip.mp4
  last frame: https://…/last.webp

With --json, the same facts are in outputMedia (index matches output).

Other useful flags:

bash
cohub generate "..." --model <model> --output ./out.png   # save locally
cohub generate "..." --model <model> --async              # queue and return
cohub generate "..." --model <model> --timeout-ms 120000  # sync wait limit
cohub generate "..." --model <model> --meta '<json>'      # pass model metadata

Speech (TTS)

Design a voice from text:

bash
cohub generate "Welcome to Cohub. This voice was created from a description." \
  --model qwen-audio-3.0-tts-plus \
  --meta '{"voice_prompt":"A calm, clear male narrator with a warm tone"}'

Clone one voice from a public reference URL:

bash
cohub generate "Welcome to Cohub. This voice follows the reference recording." \
  --model higgs-tts \
  --audio "$REFERENCE_AUDIO_URL"

Workflows

Animate a still image:

bash
cohub generate "<motion prompt>" \
  --model <model> \
  --image first_frame=./still.png

Continue a clip: pass its last frame as the next clip's first frame:

bash
cohub generate "<next shot>" \
  --model <model> \
  --image first_frame=<last frame URL>

Edit or restyle with a reference image:

bash
cohub generate "<edit instruction>" \
  --model <model> \
  --image ./input.png

For less common options, use cohub generate -h.

© netaart, 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

Files

Just SKILL.md in skills/cohub-generate of netaart/cohub.

Open the folder on GitHubat commit 1a4b7c1

Compare with similar skills

Cohub Generate 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.

Cohub Generate compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Cohub Generate this skillnetaart/cohub570—~1.2kAutomated safety check: PassApache-2.0
Media Useedenfunf/reelmimic1.7k2 repos~2kAutomated safety check: PassMIT
HyperFrames Media Useheygen-com/hyperframes59k—~2.4kAutomated safety check: PassApache-2.0
Media Genclacky-ai/openclacky1.2k—~7.3kAutomated safety check: PassMIT
Audio And Videoglifxyz/glif-mcp-server212—~1.1kAutomated safety check: PassMIT
Hyperframesscott-fryxell/brayness124—~3.2kAutomated safety check: PassMIT

Similar skills

  • Media Use

    edenfunf/reelmimic

    Agent Media OS, the single skill for every media need in a HyperFrames project.

    1.7k GitHub starsUsed in 2 repos~2k tokens
    Media & CreativeAuto-check passed
  • HyperFrames Media Use

    heygen-com/hyperframes

    Finds, generates and edits media for HyperFrames video projects: music, sound effects, images, icons, logos, voiceovers, captions and color grades.

    59k GitHub stars~2.4k tokensUpdated yesterday
    Media & CreativeAuto-check passed
  • Media Gen

    clacky-ai/openclacky

    Generate or edit images, videos, or audio in the current task.

    1.2k GitHub stars~7.3k tokensUpdated yesterday
    Media & CreativeAuto-check passed
  • Audio And Video

    glifxyz/glif-mcp-server

    Make or edit audio and video with Glif instead of writing generation code or using another tool.

    212 GitHub stars~1.1k tokensUpdated yesterday
    Media & CreativeAuto-check passed
  • Hyperframes

    scott-fryxell/brayness

    Create video compositions, animations, title cards, overlays, captions, voiceovers, audio-reactive visuals, and scene transitions in HyperFrames HTML.

    124 GitHub stars~3.2k tokensUpdated 11 days ago
    Media & CreativeAuto-check passed
  • Clean Audio

    hassancs91/claude-youtube-editor

    Voice/audio cleanup step of the AI Video Editor pipeline — diagnose a video's background noise, pick the right denoise method, and produce a cleaned master (voice isolated, levels preserved, video…

    322 GitHub starsUsed in 1 repo~1.8k tokens
    Media & CreativeAuto-check passed

More from netaart/cohub

All 29 skills in this repo
  • Pixijs Assets

    netaart/cohub

    A skill your agent uses when loading and managing resources in PixiJS v8.

    570 GitHub starsUsed in 1 repo~5k tokens
    Auto-check passed
  • A skill your agent uses when reasoning about the PixiJS v8 scene graph as a whole: how containers, leaves, transforms, and render order fit together.

    570 GitHub starsUsed in 1 repo~3.9k tokens
    Auto-check passed
  • Pixijs Scene Mesh

    netaart/cohub

    A skill your agent uses when rendering custom geometry in PixiJS v8.

    570 GitHub starsUsed in 1 repo~1.8k tokens
    Auto-check passed
  • Pixijs Scene Sprite

    netaart/cohub

    A skill your agent uses when drawing images in PixiJS v8. An agent skill from netaart/cohub.

    570 GitHub starsUsed in 1 repo~1.6k tokens
    Auto-check passed
  • Pixijs Scene Text

    netaart/cohub

    A skill your agent uses when rendering text in PixiJS v8. An agent skill from netaart/cohub.

    570 GitHub starsUsed in 1 repo~2.1k tokens
    Auto-check passed
  • A skill your agent uses when understanding how PixiJS v8 renders frames: the systems-and-pipes renderer, the render loop, and how the library adapts to different environments.

    570 GitHub stars~1.7k tokensUpdated yesterday
    Auto-check passed

Questions about Cohub Generate

What does Cohub Generate do?

Generate or transform images, video, speech, and music with Cohub multimodal models via cohub generate. Cohub Generate is an agent skill from netaart/cohub. Generate or transform images, video, speech, and music with Cohub multimodal models via cohub generate.

When should I use Cohub Generate?

Cohub Generate fits situations like: the user asks to create; remove backgrounds; synthesize speech (TTS).

How do I install Cohub Generate in Claude Code?

Run `npx skills add netaart/cohub --skill cohub-generate -a claude-code`. Or copy the skill folder (skills/cohub-generate in netaart/cohub) into .claude/skills/cohub-generate in your project. Claude Code loads it when a task matches its description.

How do I install Cohub Generate in Codex?

Run `npx skills add netaart/cohub --skill cohub-generate -a codex`. Or copy the skill folder (skills/cohub-generate in netaart/cohub) into .agents/skills/cohub-generate in your project. Codex loads it when a task matches its description.

Can I use Cohub Generate in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add netaart/cohub --skill cohub-generate -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/cohub-generate, .gemini/skills/cohub-generate, .github/skills/cohub-generate and .opencode/skills/cohub-generate in your project.

What does Cohub Generate need to run?

Going by SKILL.md and its folder, Cohub Generate needs the command-line tools its instructions call (npm). Our summary lists: Node.js.

Does Cohub Generate access the network?

SKILL.md contains no URLs. Its commands use npm, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Cohub Generate safe to install?

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. Review the folder before installing.

What licence does Cohub Generate use?

Cohub Generate 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.

How many tokens does Cohub Generate use?

About 1.2k tokens (SKILL.md is roughly 4.7k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Cohub Generate?

Skills that share tags, products or a category with Cohub Generate: Media Use (edenfunf/reelmimic, 1.7k stars), HyperFrames Media Use (heygen-com/hyperframes, 59k stars), Media Gen (clacky-ai/openclacky, 1.2k stars) and Audio And Video (glifxyz/glif-mcp-server, 212 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Cohub Generate?

netaart (a GitHub organization) maintains it in netaart/cohub, which has 570 GitHub stars. The repository holds 29 skills in this directory. The repository was last updated on October 8, 2026.

Source: netaart/cohub on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.