Whiteboard Video Workflow

Automated workflow for generating complete whiteboard animation video clips from SRT subtitle files.

Input Parameters

Parameter	Required	Description
`srtPath`	Yes	Absolute path of the SRT subtitle file
`outputDir`	No	Output root directory, defaults to the directory where the SRT file is located

Workflow Steps

The entire process is divided into 10 steps, which must be executed in strict order. Steps 3, 5, and 7 use subagent, and the remaining steps are directly executed by the main agent.

Step 0: Environment Pre-check

Run the

scripts/check_env.py

of this skill to check all dependencies at once (Python virtual environment, RUNNINGHUB_API_KEY):

bash

python3 <skill-dir>/scripts/check_env.py

Success (exit code 0): Capture
```
PYTHON_PATH=<path>
```
from the output, record this path for Step 7, continue
Failure (exit code 1): Parse each check result from the output
```
ENV_RESULT=...
```
JSON, show users clear and easy-to-understand error descriptions and repair guidance (see the troubleshooting guide below), terminate the workflow

Note: The script will automatically detect and install repairable dependencies, and only report problems that cannot be automatically repaired to the large model.

Troubleshooting Guide

When the environment pre-check fails, you must clearly explain the reason to the user and give specific repair steps. According to different failure items, give corresponding instructions:

RUNNINGHUB_API_KEY is not configured:

Whiteboard animation video generation requires RunningHub API to call AI models to generate images first. You need to configure the API Key in the
.env
file under the skill directory.
Fix method:
Add a line to the
<skill-dir>/.env
file:
RUNNINGHUB_API_KEY=your API key
If the file does not exist, create it first

The API Key can be obtained from RunningHub

Rerun after the configuration is complete

Python dependency installation failed:

Whiteboard animation relies on Python libraries such as OpenCV, NumPy, and PyAV to process videos. Automatic installation failed.

Fix method:

Please confirm that your Python version is 3.9 or higher

Then rerun this workflow, and the script will try to install automatically again

When multiple checks fail at the same time, list each failure item and corresponding repair method one by one, so that users can solve all problems at once.

Step 1: Determine Output Directory

If the user does not specify
```
outputDir
```
, use the directory where
```
srtPath
```
is located as the output root directory
Convert
```
outputDir
```
to an absolute path

Step 2: Create Output Directory Structure

Run the

scripts/workflow_helper.py

of this skill:

bash

python3 <skill-dir>/scripts/workflow_helper.py init-dirs "<outputDir>"

The output JSON contains three absolute paths:

storyboardDir

imageDir

, and

videoDir

, which are saved for later use.

Step 3: Parse SRT to Generate Storyboard Script (subagent)

Start a subagent with the instruction:

Use the Read tool to read the file
<replace this skill directory with the actual absolute path>/references/storyboard-parser.md
, and execute according to the workflow steps in it.
Input parameters:
srtPath =
<replace srtPath with the actual absolute path>
projectRoot =
<replace storyboardDir with the actual absolute path>
skill-dir =
<replace this skill directory with the actual absolute path>
(used to locate scripts)
Return the absolute path of storyboard.json and the number of scenes after completion.

Note: The main agent must fill in the actual path value into the instruction, do not pass the variable name, the subagent cannot access the context of the main agent.

You must wait for the subagent to complete and obtain the storyboard.json path before continuing.

Step 4: Parse Storyboard to Generate Image Prompts

Run the

scripts/workflow_helper.py

of this skill:

bash

python3 <skill-dir>/scripts/workflow_helper.py gen-prompts "<storyboardJsonPath>"

Output a JSON string array, each element is an image generation prompt with a whiteboard style prefix, and the array index corresponds to the order of scenes in the storyboard one by one.

At the same time, extract the

duration

value (milliseconds) of each scene from storyboard.json, and record it as an array in order for later use.

Step 5: Batch Generate Whiteboard Images (subagent)

Start a subagent with the instruction:

Use the Read tool to read the file
<replace this skill directory with the actual absolute path>/references/image-generator.md
, and execute according to the workflow steps in it.
Use batch mode, pass the following JSON string array as the prompt parameter:
<paste the actual content of the prompt JSON array output in Step 4 here>
Parameters:
skill-dir =
<replace this skill directory with the actual absolute path>
(used to locate scripts)
Output directory =
<replace imageDir with the actual absolute path>
Aspect ratio = "16:9"
Note: The main agent must fill in the actual prompt content and path value into the instruction, do not pass the variable name, the subagent cannot access the context of the main agent.

Important: Return the list of paths of all generated images, and the order must be consistent with the order of the prompt array.

You must wait for the subagent to complete and obtain all image paths before continuing.

Step 6: Verify Image Order

Confirm that the length of the image path array returned in Step 5 is consistent with the number of scenes in the storyboard, and the order is correct (the i-th image corresponds to the i-th scene).

Step 7: Batch Generate Whiteboard Animation Video Clips (subagent)

Start a subagent with the instruction:

Call the whiteboard-animation skill, use batch mode. Skip environment pre-check (the main agent has confirmed that the environment is ready), and run the batch generation script directly:
<replace PYTHON_PATH with the actual value obtained in Step 0> <replace the whiteboard-animation skill directory with the actual path>/scripts/batch_generate.py \
  --images <imagePaths[0]> <imagePaths[1]> ... \
  --durations <durations[0]> <durations[1]> ... \
  --output-dir <absolute path of videoDir>
The main agent must fill in the actual absolute path of
PYTHON_PATH
and the skill directory into the instruction, do not pass the variable name, the subagent cannot access the context of the main agent.
Parameters:
--images
: List of image paths arranged in storyboard order (separated by spaces)
--durations
: List of durations corresponding to images one by one (unit: milliseconds, directly use the duration value in the storyboard, no conversion required)
--output-dir
:
<absolute path of videoDir>
Important: After completion, collect all generated video file paths in the
<videoDir>
directory, sort them by file name timestamp, and return the complete video path list to the main agent. The order must be consistent with the order of input images.

You must wait for the subagent to complete and obtain all video paths before continuing.

Step 8: Merge Video Clips

Run the

scripts/workflow_helper.py

of this skill to merge all video clips into a complete video in order. Must use the
PYTHON_PATH
obtained in Step 0 (the PyAV dependency is in the virtual environment):

bash

<PYTHON_PATH> <skill-dir>/scripts/workflow_helper.py merge-videos "<outputDir>" <videoPath1> <videoPath2> ...

The first parameter is the output directory (the merged video is saved in this directory)
The subsequent parameters are the paths of video clips arranged in storyboard order

The output JSON contains

mergedVideo

(absolute path of the merged video),

totalSegments

, and

sizeMB

Step 9: Output Result

Output the final result, including the merged complete video path and all segment information:

Output format example:

json

{
  "mergedVideo": "/path/to/output/白板视频_20260329_120000.mp4",
  "videoSegments": [
    "/path/to/video/vid_20260329_120000_h264.mp4",
    "/path/to/video/vid_20260329_120010_h264.mp4"
  ],
  "totalSegments": 2,
  "sizeMB": 15.3,
  "outputDir": "/path/to/output"
}

Key Constraints

Step 0 must be executed before any work starts. The
```
check_env.py
```
script will automatically detect and install repairable dependencies, and only report unreparable problems to the large model
Steps 3, 5, and 7 must be executed by subagent, and the main agent waits for the result
The
```
PYTHON_PATH
```
obtained in Step 0 must be passed to the subagent of Step 7 for use, and also used for video merging in Step 8 (the PyAV dependency is in the virtual environment) to avoid repeated environment checks
Step 7 uses the batch mode of whiteboard-animation (batch_generate.py), the subagent passes in all images and durations at one time, the script generates serially internally, and returns all video paths after the subagent completes
The order of images and videos must strictly correspond to the order of scenes in the storyboard
Duration is used in milliseconds throughout the entire link, from storyboard to batch_generate.py to generate_whiteboard.py, unified as milliseconds to avoid loss of precision in floating point conversion
Step 8 uses PyAV (Python library) to merge video clips, and re-encodes through H.264 to output the final video in a unified format

Resources

references/

```
storyboard-parser.md
```
- SRT storyboard parsing workflow instruction, read and executed by the subagent in Step 3
```
image-generator.md
```
- Banana2 image generation workflow instruction, read and executed by the subagent in Step 5

scripts/

```
check_env.py
```
- One-time environment pre-check, check Python virtual environment, API Key, automatically install repairable dependencies
```
workflow_helper.py
```
- Provides three subcommands:
```
init-dirs
```
(create directory structure),
```
gen-prompts
```
(parse storyboard to generate prompts),
```
merge-videos
```
(merge video clips)
```
generate-storyboard.py
```
- Parse SRT + groups.json to generate storyboard.json
```
generate-image.py
```
- Banana2 model text-to-image, supports single and batch concurrent modes
```
banana_prompt_template.py
```
- Whiteboard style prompt template

whiteboard-video-workflow

NPX Install

Tags

SKILL.md Content (Chinese)

Whiteboard Video Workflow

Input Parameters

Workflow Steps

Step 0: Environment Pre-check

Troubleshooting Guide

Step 1: Determine Output Directory

Step 2: Create Output Directory Structure

Step 3: Parse SRT to Generate Storyboard Script (subagent)

Step 4: Parse Storyboard to Generate Image Prompts

Step 5: Batch Generate Whiteboard Images (subagent)

Step 6: Verify Image Order

Step 7: Batch Generate Whiteboard Animation Video Clips (subagent)

Step 8: Merge Video Clips

Step 9: Output Result

Key Constraints

Resources

references/

scripts/