Not in the top 1,000 by GitHub stars

Rea1Felix/H3-colab-studio

Prompt-to-Video Workspace & Google Colab G4 (ComfyUI) Agent Bridge Skill/Plugin for MiniMax H3 multi-shot video generation

History

Review weekly changes to SKILL.md before updating. Dates mark the weeks in which we observed each version; checks apply to the recorded commits.

View weekly changes

plugins/h3-colab-studio/skills/h3-colab-studio/SKILL.md

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Local scripts

First observation → 9164b8bb4e0a

No earlier version was observed for comparison.

Compare on GitHub

skills/h3-colab-studio/SKILL.md

Week of Nov 2, 2026 · Line counts unavailable

Before Not checked yet → After Local scripts

First observation → 9164b8bb4e0a

No earlier version was observed for comparison.

Compare on GitHub

View on GitHubAdd to my skills

Verified commit 9164b8bb4e0aebe578cda8cf3e4783d621254600

git clone --filter=blob:none --no-checkout https://github.com/Rea1Felix/H3-colab-studio.git .hotskills/rea1felix-h3-colab-studio && git -C .hotskills/rea1felix-h3-colab-studio checkout --detach 9164b8bb4e0aebe578cda8cf3e4783d621254600 # Copy each selected skill directory from .hotskills/rea1felix-h3-colab-studio into .claude/skills/<skill-name>/
21GitHub stars
Not availablestars in 7 days
2skills in this repository
1commit in 30 days

Heat

Insufficient history

Not enough history yet to measure heat. It appears after a few days of data.

Heat never changes the order by stars, and sponsorship never touches it. How heat works

Stars over time

Nov 7, 2026Nov 7, 202621

30-day change not yet available: not enough history.

Individual skills in this repository

How it works

h3-colab-studio

When your agent uses it

Build and operate a prompt-to-video workspace for MiniMax H3 with an agent execution bridge to Google Colab G4 running ComfyUI. Use when the user wants to generate single-shot or multi-shot MiniMax H3 videos on Colab, optimize story ideas into official H3 structured prompts, build a job-queue control app for H3 video production, or stitch and verify multi-shot H3 clips.

What it loads into context

Frontmatter

105 tokens

Always available to the agent

SKILL.md

1,802 tokens

When the skill is activated

Reference documents

3,247 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

  1. What this builds
  2. Set up Colab and H3
  3. Build the workspace app
  4. Write the H3 prompts
  5. Execute a job and finish it
  6. Measured reference points
  7. Pitfalls
  8. Browser & connector integration
  9. Websites

Scripts executed separately

  • scripts/stitch.sh

Files

  • ../../../../.claude-plugin/marketplace.jsonasset382 bytes
  • ../../../../.claude-plugin/plugin.jsonasset597 bytes
  • SKILL.mdcore8,053 bytes
  • SKILL.mdcore8,053 bytes
  • references/app-blueprint.mddoc3,320 bytes
  • references/colab-runbook.mddoc5,004 bytes
  • references/prompt-spec.mddoc4,736 bytes
  • scripts/stitch.shruns1,099 bytes
  • skill.yamlasset428 bytes
Read SKILL.md

# H3 Colab Studio — Prompt-to-Video Workspace with an Agent Bridge

## What this builds

- Build three cooperating parts: a job-queue app the user drives, a prompt optimizer that rewrites ideas into the official MiniMax H3 prompt structure, and yourself as the execution bridge to Google Colab.

- Treat the app as the control surface only. GPU generation happens on Colab, never inside the app, and the app must say so plainly instead of faking a live generate button.

- Expect the payoff to be repeatable multi-shot videos. A single H3 generation is one continuous take with no cuts, so story structure comes from generating one clip per shot and stitching, which this workflow makes routine.

## Set up Colab and H3

- Confirm the user has a Google account with Colab compute units available, for example through Google AI Ultra or a Colab paid plan, and warn them that granted unit pools can expire (the Ultra grant lasts about 90 days).
View the complete source on GitHub

Skills included

  • h3-colab-studio
  • h3-colab-studio

Repository facts

License
MIT
Created
Nov 6, 2026
Last pushed
Nov 6, 2026
Forks
5
Open issues
0
Project activity
Active

Before you install

Local scripts checked with an earlier scanner version

Automated signals, not a guarantee

Checked Nov 8, 2026

Not fully analyzed: some files were too long, unreadable or missing. This never lowers the level.

Show 2 notes
  • plugins/h3-colab-studio/skills/h3-colab-studio/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • skills/h3-colab-studio/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
Show all 18 signals by file

.claude-plugin/marketplace.json

  • Network accessline 3
    [Credential-related source omitted]
  • Network accessline 4
    [Credential-related source omitted]
  • Network accessline 5
    [Credential-related source omitted]

.claude-plugin/plugin.json

  • Network accessline 5
    [Credential-related source omitted]
  • Network accessline 6
    [Credential-related source omitted]
  • Network accessline 7
    [Credential-related source omitted]
  • Network accessline 8
    [Credential-related source omitted]
  • Network accessline 9
    [Credential-related source omitted]

plugins/h3-colab-studio/skills/h3-colab-studio/SKILL.md

  • Network accessline 97
    [Credential-related source omitted]
  • Network accessline 99
    [Credential-related source omitted]
  • Network accessline 101
    [Credential-related source omitted]

plugins/h3-colab-studio/skills/h3-colab-studio/scripts/stitch.sh

  • Executable scriptline 1
    [Credential-related source omitted]
  • Dangerous shell operationline 15
    [Credential-related source omitted]

skills/h3-colab-studio/SKILL.md

  • Network accessline 97
    [Credential-related source omitted]
  • Network accessline 99
    [Credential-related source omitted]
  • Network accessline 101
    [Credential-related source omitted]

skills/h3-colab-studio/scripts/stitch.sh

  • Executable scriptline 1
    [Credential-related source omitted]
  • Dangerous shell operationline 15
    [Credential-related source omitted]

How the signals fit together

Facts from a static read of the files at the pinned commit. They describe what the code does, not why.

Some files could not be read; the facts below still hold.

Signal confidence: 10 facts

Works with

Agent compatibility, from the files in this repository
AgentStatusEvidenceInstall to
Claude CodeNativeManifest .claude-plugin/marketplace.json.claude/skills/<skill-name>/
CodexCompatibleStandard SKILL.md plugins/h3-colab-studio/skills/h3-colab-studio/SKILL.md.agents/skills/<skill-name>/
CursorCompatibleStandard SKILL.md plugins/h3-colab-studio/skills/h3-colab-studio/SKILL.md.cursor/skills/<skill-name>/
Gemini CLICompatibleStandard SKILL.md plugins/h3-colab-studio/skills/h3-colab-studio/SKILL.md.gemini/skills/<skill-name>/
GitHub CopilotCompatibleStandard SKILL.md plugins/h3-colab-studio/skills/h3-colab-studio/SKILL.md.github/skills/<skill-name>/
OpenCodeCompatibleStandard SKILL.md plugins/h3-colab-studio/skills/h3-colab-studio/SKILL.md.opencode/skills/<skill-name>/
  • Native The repository ships a manifest, folder or frontmatter field for this agent.
  • Compatible A standard SKILL.md that this agent can read.
  • Unknown No evidence either way.

Derived automatically from paths, manifests and frontmatter. Not tested on each agent.