Seedance 2.0 Storyboard Creation & Video Generation

End-to-end workflow from idea to final video: Storyboard Guidance → Reference Image Generation → Video Submission → Result Retrieval.

Step 0: Determine Execution Mode (MCP or Script)

Prioritize checking MCP availability:

Check the status of the
```
xskill-ai
```
MCP service (read
```
mcps/user-xskill-ai/STATUS.md
```
)
If MCP is available → Use MCP tools like
```
submit_task
```
/
```
get_task
```
If MCP is unavailable or errors occur → Switch to Script Mode

Prerequisites for Script Mode:

Confirm that the environment variable
```
XSKILL_API_KEY
```
is set (execute
```
echo $XSKILL_API_KEY | head -c 10
```
in Shell)

If not set, prompt the user:

export XSKILL_API_KEY=sk-your-api-key
Get API Key: https://www.xskill.ai/#/v2/api-keys

Confirm
```
requests
```
is installed (
```
pip install requests
```
)

Script Path:

scripts/seedance_api.py

in this Skill directory, locate it via:

bash

# Search via Glob tool
glob: .cursor/skills/seedance2-api/scripts/seedance_api.py

In subsequent steps, each API call provides both MCP Method and Script Method, choose one based on the judgment from Step 0.

Step 1: Understand User's Idea

Collect the following information (proactively ask if missing):

Core Story: One-sentence summary of what the video is about
Duration: 4-15 seconds
Aspect Ratio: 16:9 / 9:16 / 1:1 / 21:9 / 4:3 / 3:4
Visual Style: Realistic/Animated/Ink Wash/Sci-Fi/Cyberpunk, etc.
Material Status: Whether there are existing images/videos/audio, or if AI generation is needed
Function Mode: Whether first-last frame control (
```
first_last_frames
```
) is required; otherwise, default to omni-reference mode (
```
omni_reference
```
)

Step 2: Deep Dive (5 Dimensions)

Guide the user to supplement details for each dimension:

Content - Who is the protagonist? What are they doing? Where is it set?
Visuals - Lighting, color tone, texture, atmosphere
Camera - Push/Pull/Pan/Truck/Follow/Orbit/Crane
Action - Specific actions and rhythm of the subject
Audio - Soundtrack style, sound effects, dialogue

Step 3: Build Storyboard Structure

Split shots by timeline using the following formula:

【Style】_____ style, _____ seconds, _____ aspect ratio, _____ atmosphere

0-X seconds: [Camera Movement] + [Frame Content] + [Action Description]
X-Y seconds: [Camera Movement] + [Frame Content] + [Action Description]
...

【Audio】_____ soundtrack + _____ sound effects + _____ dialogue
【Reference】@image_file_1 _____, @video_file_1 _____

Detailed templates and examples can be found in reference.md.

Step 4: Generate Reference Images (If Needed)

If the user has no existing materials, use Seedream 4.5 to generate character images, scene images, first/last frames, etc.

Text-to-Image

<details> <summary>MCP Method</summary>

Call the

submit_task

tool:

model_id:

fal-ai/bytedance/seedream/v4.5/text-to-image

parameters:
- prompt: Detailed image description (English works better)
- image_size: Select based on video aspect ratio
- num_images: Number of images needed (1-6)

</details> <details> <summary>Script Method</summary>

bash

python .cursor/skills/seedance2-api/scripts/seedance_api.py submit \
  --model "fal-ai/bytedance/seedream/v4.5/text-to-image" \
  --params '{"prompt":"An astronaut in a white spacesuit...","image_size":"landscape_16_9","num_images":1}'

</details>

Image Editing (Modify Based on Existing Images)

<details> <summary>MCP Method</summary>

Call the

submit_task

tool:

model_id:
```
fal-ai/bytedance/seedream/v4.5/edit
```
parameters:
- prompt: Editing instructions (refer to images using Figure 1/2/3)
- image_urls: Array of input image URLs
- image_size: Output size

</details> <details> <summary>Script Method</summary>

bash

python .cursor/skills/seedance2-api/scripts/seedance_api.py submit \
  --model "fal-ai/bytedance/seedream/v4.5/edit" \
  --params '{"prompt":"Change the background to a forest","image_urls":["https://..."],"image_size":"landscape_16_9"}'

</details>

Poll for Image Results

Images take approximately 1-2 minutes to complete.

<details> <summary>MCP Method</summary>

Call the

get_task

tool to query status:

First query after 30 seconds
Then query every 30 seconds thereafter
Extract image URLs when status is
```
completed
```

</details> <details> <summary>Script Method</summary>

Single Query:

bash

python .cursor/skills/seedance2-api/scripts/seedance_api.py query \
  --task-id "TASK_ID_HERE"

Auto Polling (Recommended for Images, 10s interval, 180s timeout):

bash

python .cursor/skills/seedance2-api/scripts/seedance_api.py poll \
  --task-id "TASK_ID_HERE" --interval 10 --timeout 180

</details>

image_size Reference Table

Video Aspect Ratio	Recommended image_size	Description
16:9	landscape_16_9	Landscape
9:16	portrait_16_9	Portrait
4:3	landscape_4_3	Landscape
3:4	portrait_4_3	Portrait
1:1	square_hd	Square
21:9	landscape_16_9	Approximate ultra-wide screen

Step 5: Generate Professional Prompts

Integrate the storyboard structure and reference images into the final prompt:

Use
```
@image_file_1
```
,
```
@image_file_2
```
, etc. to reference images in the
```
image_files
```
array
Use
```
@video_file_1
```
, etc. to reference videos in the
```
video_files
```
array
Use
```
@audio_file_1
```
, etc. to reference audio in the
```
audio_files
```
array
Also compatible with the old syntax:
```
@图片1
```
,
```
@视频1
```
,
```
@音频1
```

Reference Syntax Example:

@image_file_1 as character reference, refer to the camera movement of @video_file_1, match the soundtrack of @audio_file_1

Important: The Nth URL in the

image_files

array corresponds to

@image_file_N

video_files

and

audio_files

are numbered independently.

Step 6: Submit Video Task

Process Material URLs:

Images generated by Seedream: URLs already exist, use directly
Web images provided by user: Use directly
Local images provided by user: Upload first to get URL (see upload methods below)

Upload Local Images

<details> <summary>MCP Method</summary>

Call the

upload_image

tool: image_url or image_data

</details> <details> <summary>Script Method</summary>

bash

# Upload web image
python .cursor/skills/seedance2-api/scripts/seedance_api.py upload \
  --image-url "https://example.com/image.png"

# Upload local image
python .cursor/skills/seedance2-api/scripts/seedance_api.py upload \
  --image-path "/path/to/local/image.png"

</details>

Submit Seedance 2.0 Task (Omni-reference Mode

omni_reference

)

<details> <summary>MCP Method</summary>

Call the

submit_task

tool:

model_id:
```
st-ai/super-seed2
```
parameters:
- prompt: Complete prompt generated in Step 5
- functionMode:
```
omni_reference
```
  (default, can be omitted)
- image_files: Array of reference image URLs (max 9 images, order corresponds to @image_file_1/2/3...)
- video_files: Array of reference video URLs (max 3, total duration ≤ 15s)
- audio_files: Array of reference audio URLs (max 3)
- ratio: Aspect ratio (
```
16:9
```
  /
```
9:16
```
  /
```
1:1
```
  /
```
21:9
```
  /
```
4:3
```
  /
```
3:4
```
  )
- duration: Integer duration (4 - 15)
- model:
```
seedance_2.0_fast
```
  (default fast mode) or
```
seedance_2.0
```
  (standard quality)

</details> <details> <summary>Script Method</summary>

bash

python .cursor/skills/seedance2-api/scripts/seedance_api.py submit \
  --model "st-ai/super-seed2" \
  --params '{
    "prompt": "Cinematic realistic sci-fi style, 15 seconds, 16:9...",
    "functionMode": "omni_reference",
    "image_files": ["https://img1.png", "https://img2.png"],
    "ratio": "16:9",
    "duration": 15,
    "model": "seedance_2.0_fast"
  }'

</details>

Submit Seedance 2.0 Task (First-Last Frame Mode

first_last_frames

)

<details> <summary>MCP Method</summary>

Call the

submit_task

tool:

model_id:
```
st-ai/super-seed2
```
parameters:
- prompt: Video description prompt
- functionMode:
```
first_last_frames
```
- filePaths: Array of image URLs (0 images = text-to-video, 1 image = first-frame-to-video, 2 images = first-last-frame video)
- ratio: Aspect ratio
- duration: Integer duration
- model:
```
seedance_2.0_fast
```
  or
```
seedance_2.0
```

</details> <details> <summary>Script Method</summary>

bash

python .cursor/skills/seedance2-api/scripts/seedance_api.py submit \
  --model "st-ai/super-seed2" \
  --params '{
    "prompt": "Camera transitions smoothly from first frame to last frame, with natural and fluid visuals",
    "functionMode": "first_last_frames",
    "filePaths": ["https://first-frame.png", "https://last-frame.png"],
    "ratio": "16:9",
    "duration": 5,
    "model": "seedance_2.0_fast"
  }'

</details>

Step 7: Poll for Video Results

Video generation takes approximately 10 minutes.

<details> <summary>MCP Method</summary>

Polling Strategy:

After submission, inform the user: "Video is being generated, expected to take 10 minutes"
First call
```
get_task
```
to query after 60 seconds
Then query every 90 seconds thereafter
Report the status to the user after each query

</details> <details> <summary>Script Method</summary>

Recommended: Use auto polling (runs in background, 30s interval, 600s timeout):

bash

python .cursor/skills/seedance2-api/scripts/seedance_api.py poll \
  --task-id "TASK_ID_HERE" --interval 30 --timeout 600

The script will print polling progress to stderr, and output JSON results to stdout upon completion.

Manual Query in Batches:

bash

python .cursor/skills/seedance2-api/scripts/seedance_api.py query \
  --task-id "TASK_ID_HERE"

</details>

Status Explanations:

```
pending
```
→ "Queued..."
```
processing
```
→ "Generating..."
```
completed
```
→ Extract video URL and present to user
```
failed
```
→ Inform user of failure reason, suggest adjusting prompt and retrying

Complete Workflow Example

User says: "Help me make a video of an astronaut walking on Mars"

When MCP is Available

1. Collect information → 15 seconds, 16:9, cinematic sci-fi style, no existing materials

2. Generate astronaut image + Mars scene image using Seedream 4.5
   submit_task("fal-ai/bytedance/seedream/v4.5/text-to-image", {...})
   → Poll get_task → Obtain image URLs

3. Write prompt → Submit video task
   submit_task("st-ai/super-seed2", {...})

4. Poll get_task, obtain video URL after ~10 minutes

When MCP is Unavailable (Script Mode)

1. Collect information → 15 seconds, 16:9, cinematic sci-fi style

2. Generate reference images:
   python scripts/seedance_api.py submit \
     --model "fal-ai/bytedance/seedream/v4.5/text-to-image" \
     --params '{"prompt":"An astronaut in white spacesuit on Mars...","image_size":"landscape_16_9"}'
   → Get task_id

3. Poll for image results:
   python scripts/seedance_api.py poll --task-id "xxx" --interval 10 --timeout 180
   → Get image URLs

4. Submit video task:
   python scripts/seedance_api.py submit \
     --model "st-ai/super-seed2" \
     --params '{"prompt":"...storyboard prompt...","functionMode":"omni_reference","image_files":["image URL"],"ratio":"16:9","duration":15,"model":"seedance_2.0_fast"}'
   → Get task_id

5. Poll for video results:
   python scripts/seedance_api.py poll --task-id "xxx" --interval 30 --timeout 600
   → Get video URL

Quick Reference for Model Parameters

Seedream 4.5 Text-to-Image

Parameter	Type	Required	Description
prompt	string	Yes	Image description
image_size	string	No	auto_2K / auto_4K / square_hd / portrait_4_3 / portrait_16_9 / landscape_4_3 / landscape_16_9
num_images	int	No	1-6, default 1

Seedream 4.5 Image Editing

Parameter	Type	Required	Description
prompt	string	Yes	Editing instructions, reference with Figure 1/2/3
image_urls	array	Yes	List of input image URLs
image_size	string	No	Same as above
num_images	int	No	1-6, default 1

Seedance 2.0 Video (Omni-reference Mode

omni_reference

)

Parameter	Type	Required	Description
prompt	string	Yes	Storyboard prompt, reference with @image_file_N/@video_file_N/@audio_file_N
functionMode	string	No	`omni_reference` (default)
image_files	array	No	Array of reference image URLs (max 9)
video_files	array	No	Array of reference video URLs (max 3, total duration ≤ 15s)
audio_files	array	No	Array of reference audio URLs (max 3)
ratio	string	No	21:9 / 16:9 / 4:3 / 1:1 / 3:4 / 9:16
duration	integer	No	4-15, default 5
model	string	No	seedance_2.0_fast (default) / seedance_2.0

Seedance 2.0 Video (First-Last Frame Mode

first_last_frames

)

Parameter	Type	Required	Description
prompt	string	Yes	Video description prompt
functionMode	string	Yes	`first_last_frames`
filePaths	array	No	Array of image URLs (0 = text-to-video, 1 = first frame, 2 = first-last frames)
ratio	string	No	21:9 / 16:9 / 4:3 / 1:1 / 3:4 / 9:16
duration	integer	No	4-15, default 5
model	string	No	seedance_2.0_fast (default) / seedance_2.0

Quick Reference for Tools

MCP Tools

Operation	Tool	Key Parameters
Submit Task	submit_task	model_id, parameters
Query Result	get_task	task_id
Upload Image	upload_image	image_url or image_data
Query Balance	get_balance	None

Script Commands (When MCP is Unavailable)

Operation	Command	Description
Submit Task	`python scripts/seedance_api.py submit --model MODEL --params '{...}'`	Returns task_id
Single Query	`python scripts/seedance_api.py query --task-id ID`	Returns current status
Auto Polling	`python scripts/seedance_api.py poll --task-id ID --interval N --timeout N`	Blocks until completion
Query Balance	`python scripts/seedance_api.py balance`	Returns account balance
Upload Image	`python scripts/seedance_api.py upload --image-url URL` or `--image-path PATH`	Returns image URL

Script Path Note: The
scripts/seedance_api.py
in the above commands is relative to the
.cursor/skills/seedance2-api/
directory. When executing, use the full path
.cursor/skills/seedance2-api/scripts/seedance_api.py
, or cd to the skill directory first.

Seedance 2.0 Limitations

Does not support uploading realistic human face materials
Maximum 12 files: images ≤9 + videos ≤3 + audio ≤3
Total duration of video/audio references ≤15 seconds
Using video references consumes more credits

Additional Resources

Detailed storyboard templates, complete examples, and camera vocabulary can be found in reference.md.

seedance2-api

NPX Install

Tags

SKILL.md Content (Chinese)

Seedance 2.0 Storyboard Creation & Video Generation

Step 0: Determine Execution Mode (MCP or Script)

Step 1: Understand User's Idea

Step 2: Deep Dive (5 Dimensions)

Step 3: Build Storyboard Structure

Step 4: Generate Reference Images (If Needed)

Text-to-Image

Image Editing (Modify Based on Existing Images)

Poll for Image Results

image_size Reference Table

Step 5: Generate Professional Prompts

Step 6: Submit Video Task

Upload Local Images

Submit Seedance 2.0 Task (Omni-reference Mode
`omni_reference`
)

Submit Seedance 2.0 Task (First-Last Frame Mode
`first_last_frames`
)

Step 7: Poll for Video Results

Complete Workflow Example

When MCP is Available

When MCP is Unavailable (Script Mode)

Quick Reference for Model Parameters

Seedream 4.5 Text-to-Image

Seedream 4.5 Image Editing

Seedance 2.0 Video (Omni-reference Mode
`omni_reference`
)

Seedance 2.0 Video (First-Last Frame Mode
`first_last_frames`
)

Quick Reference for Tools

MCP Tools

Script Commands (When MCP is Unavailable)

Seedance 2.0 Limitations

Additional Resources

seedance2-api

NPX Install

Tags

SKILL.md Content (Chinese)

Seedance 2.0 Storyboard Creation & Video Generation

Step 0: Determine Execution Mode (MCP or Script)

Step 1: Understand User's Idea

Step 2: Deep Dive (5 Dimensions)

Step 3: Build Storyboard Structure

Step 4: Generate Reference Images (If Needed)

Text-to-Image

Image Editing (Modify Based on Existing Images)

Poll for Image Results

image_size Reference Table

Step 5: Generate Professional Prompts

Step 6: Submit Video Task

Upload Local Images

Submit Seedance 2.0 Task (Omni-reference Mode omni_reference)

Submit Seedance 2.0 Task (First-Last Frame Mode first_last_frames)

Step 7: Poll for Video Results

Complete Workflow Example

When MCP is Available

When MCP is Unavailable (Script Mode)

Quick Reference for Model Parameters

Seedream 4.5 Text-to-Image

Seedream 4.5 Image Editing

Seedance 2.0 Video (Omni-reference Mode omni_reference)

Seedance 2.0 Video (First-Last Frame Mode first_last_frames)

Quick Reference for Tools

MCP Tools

Script Commands (When MCP is Unavailable)

Seedance 2.0 Limitations

Additional Resources

Submit Seedance 2.0 Task (Omni-reference Mode
`omni_reference`
)

Submit Seedance 2.0 Task (First-Last Frame Mode
`first_last_frames`
)

Seedance 2.0 Video (Omni-reference Mode
`omni_reference`
)

Seedance 2.0 Video (First-Last Frame Mode
`first_last_frames`
)