AgentMatter
EN/中文
SubmitGitHub
Skills · Community project

SRT Whiteboard Animation

SRT Whiteboard Animation parses subtitles, plans scenes, coordinates line art and semantic regions, then renders continuous ink-and-color whiteboard animation to MP4.

2.3k StarsMITPythonUpdated 1 months ago
Install and use GitHub
01

Project overview

SRT Whiteboard Animation is an Agent Skill for subtitle-led video production. It parses SRT files into suggested 25–35 second scenes, pauses for approval of storyboards, line art, and pixel regions, then renders each element as continuous ink followed by restrained color. It is aimed at explainers, lessons, narrated stories, and short-form scripts rather than one-click full-frame reveal effects.

Monkey Mountain whiteboard illustration example from SRT Whiteboard Animation
The repository example shows a warm paper background, dark sketch lines, and restrained color accents.View repository image
02

Core capabilities

01

Subtitle-led storyboarding

parse_srt.py parses subtitles and proposes scenes around the documented target duration.

02

Semantic region annotation

Visible subjects are linked to subtitle events through sequence, startMs, and protectedRegions.

03

Continuous-stroke rendering

Each region follows a grid or skeleton path, drawing ink first and color second.

04

Multi-scene merge

merge_scenes.py combines approved scene videos in a specified order.

03

Installation and usage

Download the repository, check or create its isolated Python environment with prepare_env.py, and use the reported ENV_PY interpreter for rendering.

AI AGENT INSTALL

Let an AI Agent install it

Send this prompt to Codex, Claude Code, or another AI agent that can work with your local environment.

Install the SRT Whiteboard Animation Skill from https://github.com/geeklee/srt-whiteboard-animation. Read README.md, SKILL.md, and scripts/prepare_env.py first. Download it into the project- or user-level Skills directory I specify, run `python scripts/prepare_env.py --check`, and only run `python scripts/prepare_env.py` if dependencies are missing. Verify with my SRT using parse_srt.py and report ENV_PY and the proposed scenes. Ask before installing dependencies, opening a browser, requesting file access, overwriting files, or running any additional command.
01Before you start
  • A working Python command
  • Chrome or Edge for the local preview UI
  • An SRT file and later-approved scene images
  • Write access and network access for first-time dependencies
02Copy the install command or configuration
Project URL: https://github.com/geeklee/srt-whiteboard-animation
Copy the URL and follow the steps below to complete setup.
03Complete the setup steps
  1. 1
    Prepare the environment

    Run `python scripts/prepare_env.py --check`; if missing, run `python scripts/prepare_env.py` and capture ENV_PY.

  2. 2
    Parse subtitles and approve scenes

    Run `python scripts/parse_srt.py <subtitles.srt> --target-sec 30 --min-sec 25 --max-sec 35`, review the proposed scenes, and stop for approval.

  3. 3
    Annotate and preview

    Create the matching annotation.json, open assets/preview.html, and inspect regions, order, timing, and subtitle links.

  4. 4
    Render and merge

    Render each approved scene with render_stream_whiteboard.py, then merge only after every scene is accepted.

How to verify the setup

Run `python scripts/prepare_env.py --check` from the repository root and confirm the final line prints `ENV_PY=<path>`. Then run parse_srt.py on a sample SRT and confirm it reports subtitle entries, suggested scenes, and sceneDurationMs.

Before using it
  • Use the ENV_PY interpreter for later render commands.
  • SKILL.md requires explicit user approval after each major stage.
  • Writing annotations from the preview UI requires Chrome or Edge File System Access API support.
04

Use cases

SCENARIO 01

Educational explainers

Draw concepts and objects in the same order as a lesson or science narration.

SCENARIO 02

Narrated stories

Tie characters, actions, and outcomes to subtitle events so later elements stay hidden.

SCENARIO 03

Short-form whiteboard packaging

Produce consistent warm-paper MP4 scenes for downstream editing.

05

Assessment

The public README, SKILL.md, scripts, and examples show a concrete subtitle-to-render workflow with clear timing and mask rules. The tradeoff is a multi-stage process that depends on user approval and accurate pixel regions. This is a documentation-based assessment, not an installation or rendering test.

Why it may be useful

  • Complete path from subtitle parsing to scene merging
  • Approval gates reduce expensive downstream rework
  • Includes examples, a preview UI, and quality checks

What to know first

  • Requires Python dependencies and browser file access
  • Accurate regions still need visual review
  • Not designed for one-click photorealistic or 3D animation
06

README

SRT Whiteboard Animation

Overview

SRT Whiteboard Animation parses subtitles, plans scenes, coordinates line art and semantic regions, then renders continuous ink-and-color whiteboard animation to MP4. Turn SRT subtitles into warm paper whiteboard animation with regional mask orchestration and streaming ink-to-color strokes.

Getting started

  • Download the repository, check or create its isolated Python environment with prepare_env.py, and use the reported ENV_PY interpreter for rendering.
  • The parser proposes scenes from subtitle timing, with one core idea per scene.
  • annotation.json stores regions, subtitle links, sequence, timing, and overlap protection.
  • Run `python scripts/prepare_env.py --check` from the repository root and confirm the final line prints `ENV_PY=<path>`. Then run parse_srt.py on a sample SRT and confirm it reports subtitle entries, suggested scenes, and sceneDurationMs.

Configuration

Project URL: https://github.com/geeklee/srt-whiteboard-animation
Copy the URL and follow the steps below to complete setup.
Read the complete README on GitHub →