Skill: Seedance 2 Video Generation

Owner: EvoLinkAI

Summary: Seedance 2.0 AI video generation via EvoLink API. Text-to-video, image-to-video with auto audio (voice, SFX, BGM). Works with OpenClaw, Claude Code, Cursor....

Tags: latest:1.4.1

Version history:

v1.4.1 | 2026-02-23T18:02:09.524Z | user

Fix: silent polling to prevent message spam during generation

v1.4.0 | 2026-02-23T17:43:03.552Z | user

UX: add post-install greeting flow, soften prompt guidance to allow offering inspiration

v1.3.1 | 2026-02-23T17:38:06.682Z | user

SEO: optimize description keywords

v1.3.0 | 2026-02-23T17:26:18.244Z | user

Security hardening: fix metadata format (env as string array + primaryEnv), add homepage for provenance, replace all regex JSON parsing with jq, hardcode API endpoint as readonly

v1.2.1 | 2026-02-23T17:23:01.673Z | user

Fix name for better search discoverability

v1.2.0 | 2026-02-23T17:06:07.432Z | user

Security fix: declare EVOLINK_API_KEY env + jq/curl bin dependencies in metadata; fix duration inconsistency (4-12s everywhere)

v1.1.0 | 2026-02-23T17:00:03.552Z | user

Interactive guided flow: smart intent detection, no default assumptions, user decides all parameters

v1.0.1 | 2026-02-23T16:14:10.322Z | user

Fix: enable audio by default, correct duration 4-12s, max 2 images, add --no-audio flag

v1.0.0 | 2026-02-23T15:56:14.132Z | user

Initial release: text-to-video, image-to-video, multi-modal reference support, auto-polling, friendly error handling

Archive index:

Archive v1.4.1: 4 files, 8991 bytes

Files: _meta.json (139b), references/api-params.md (6220b), scripts/seedance-gen.sh (9004b), SKILL.md (5776b)

File v1.4.1:SKILL.md

---

name: seedance-2-video-gen description: Seedance 2.0 AI video generation via EvoLink API. Text-to-video, image-to-video with auto audio (voice, SFX, BGM). Works with OpenClaw, Claude Code, Cursor. Powered by ByteDance Seedance model. homepage: https://github.com/EvoLinkAI/evolink-skills metadata: {"openclaw":{"homepage":"https://github.com/EvoLinkAI/evolink-skills","requires":{"bins":["jq","curl"],"env":["EVOLINK_API_KEY"]},"primaryEnv":"EVOLINK_API_KEY"}}

Seedance Video Generation

An interactive AI video generation assistant powered by the Seedance model via EvoLink API.

After Installation

When this skill is first loaded, proactively greet the user and start the setup:

1. Check if EVOLINK_API_KEY is set

   • If not set: "To generate videos, you'll need an EvoLink API key. It takes 30 seconds to get one — just sign up at evolink.ai. Want me to walk you through it?"
   • If already set: "You're all set! What kind of video would you like to create?"

2. That's it. One question. The user is now in the flow.

Do NOT list features, show a menu, or dump instructions. Just ask one question to move forward.

Core Principles

1. Guide, don't decide — Present options and let the user decide. Don't make assumptions about their preferences.
2. Let the user drive the creative vision — If they have an idea, use their words. If they need inspiration, offer suggestions and let them choose or refine.
3. Smart context awareness — Recognize what the user has already provided and only ask about missing pieces.
4. Intent first — If the user's intent is unclear, confirm what they want before proceeding.

Flow

Step 1: Check for API Key

If the user hasn't provided an API key or set EVOLINK_API_KEY:

• Tell them they need an EvoLink API Key
• Guide them to register at https://evolink.ai and get a key from the dashboard
• Once they provide a key, proceed to Step 2

If the key is already set or provided, skip directly to Step 2.

Step 2: Understand Intent

Assess what the user wants based on their message:

• Intent is clear (e.g., "generate a video of a cat dancing") → Go to Step 3
• Intent is ambiguous (e.g., "I want to try Seedance") → Ask what they'd like to do: generate a new video, learn about model capabilities, etc.

Step 3: Gather Missing Information

Check what the user has already provided and only ask about what's missing:

| Parameter | What to tell the user | Required? | |-----------|----------------------|-----------| | Video content (prompt) | Ask what they'd like to see. If they need inspiration, suggest a few ideas for them to pick from or build on. | Yes | | Duration | Supported: 4–12 seconds. Ask how long they want. | Yes | | Resolution | Supported: 480p / 720p / 1080p. Ask their preference. | Yes | | Audio | The model can auto-generate voice, sound effects, and background music matching the video. Ask if they want audio enabled. | Yes | | Aspect ratio | Supported: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9. Only mention if relevant or if user asks. | Optional | | Reference images | Supported: up to 9 images (JPEG/PNG/WebP, ≤30MB each). Only mention if relevant. | Optional |

Smart gathering rules:

• User gives everything at once → Confirm and generate immediately
• User gives partial info → Only ask about the missing pieces
• User says "I want to generate a video" with no details → Guide from the beginning

Step 4: Generate

Once all required information is confirmed:

1. Tell the user: "Generating your video now — this usually takes 30–120 seconds. I'll let you know when it's ready."
2. Run the generation script. Do NOT forward each line of script output to the user. The script prints polling status internally — ignore it. Only report the final result.
3. When complete, share the video URL (valid for 24 hours) and generation time.

Script Usage

# Set API key
export EVOLINK_API_KEY=your_key_here

# Basic text-to-video
./scripts/seedance-gen.sh "user's prompt" --duration 5 --quality 720p

# With audio disabled
./scripts/seedance-gen.sh "user's prompt" --duration 8 --quality 1080p --no-audio

# With reference image
./scripts/seedance-gen.sh "user's prompt" --image "https://example.com/ref.jpg" --duration 6 --quality 720p

# Custom aspect ratio
./scripts/seedance-gen.sh "user's prompt" --aspect-ratio 9:16 --duration 4 --quality 480p

Error Handling

Provide friendly, actionable messages:

| Error | What to tell the user | |-------|----------------------| | Invalid/missing key (401) | "Your API key doesn't seem to work. You can check it at https://evolink.ai/dashboard" | | Insufficient balance (402) | "Your account balance is low. You can add credits at https://evolink.ai/dashboard" | | Rate limited (429) | "Too many requests — let's wait a moment and try again" | | Content blocked (400) | "This prompt was flagged (realistic human faces are restricted). Try adjusting the description" | | Service unavailable (503) | "The service is temporarily busy. Let's try again in a minute" |

Model Capabilities Summary

Use this when the user asks what the model can do:

• Text-to-video: Describe a scene, get a video
• Image-to-video: Provide reference images to guide the output
• Audio generation: Auto-generates synchronized voice, sound effects, and background music
• Duration: 4–12 seconds
• Resolution: 480p, 720p, 1080p
• Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9
• Limitation: Realistic human faces are restricted

References

• references/api-params.md: Complete API parameter reference
• scripts/seedance-gen.sh: Generation script with automatic polling and error handling

File v1.4.1:_meta.json

{ "ownerId": "kn74p4xy6sja0199cea53anecs81kqjs", "slug": "seedance-2-video-gen", "version": "1.4.1", "publishedAt": 1771869729524 }

File v1.4.1:references/api-params.md

Seedance API Parameters Reference

This document provides complete API parameter reference for the Seedance video generation service.

API Endpoints

Generation Request

POST https://api.evolink.ai/v1/videos/generations
Authorization: Bearer {EVOLINK_API_KEY}
Content-Type: application/json

Task Status Query

GET https://api.evolink.ai/v1/tasks/{task_id}
Authorization: Bearer {EVOLINK_API_KEY}

Generation Parameters

Required Parameters

| Parameter | Type | Description | |-----------|------|-------------| | model | string | Fixed: "seedance-1.5-pro" | | prompt | string | Text description of desired video content |

Optional Parameters

| Parameter | Type | Default | Options | Description | |-----------|------|---------|---------|-------------| | image_urls | array | [] | URLs | Reference images (max 9) | | duration | integer | 5 | 4-12 | Video duration in seconds | | quality | string | "720p" | "480p", "720p", "1080p" | Video resolution | | aspect_ratio | string | "16:9" | "16:9", "9:16", "1:1", "4:3", "3:4", "21:9" | Video aspect ratio | | generate_audio | boolean | true | true, false | Generate synchronized audio (voice, SFX, music). Enabled by default; disabling reduces cost. |

Parameter Constraints

Image URLs

• Maximum count: 9 images
• File size limit: ≤30MB per image
• Supported formats: JPEG, PNG, WebP, BMP, TIFF, GIF
• Format: Array of strings containing full URLs

Duration

• Range: 4-12 seconds
• Type: Integer
• Impact: Longer videos cost more and take longer to generate

Quality

• 480p: Fastest, cheapest, lower quality
• 720p: Balanced (recommended default)
• 1080p: Highest quality, slower, most expensive

Aspect Ratio

• 16:9: Standard landscape (YouTube, TV)
• 9:16: Mobile/vertical (TikTok, Instagram Stories)
• 1:1: Square (Instagram posts, social media)
• 4:3: Traditional format
• 3:4: Portrait format
• 21:9: Ultra-wide cinematic

Example Payloads

Text-to-Video

{
  "model": "seedance-1.5-pro",
  "prompt": "A serene sunset over calm ocean waves",
  "duration": 5,
  "quality": "720p",
  "aspect_ratio": "16:9",
  "generate_audio": false
}

Image-to-Video

{
  "model": "seedance-1.5-pro", 
  "prompt": "Gentle waves rolling onto the shore",
  "image_urls": [
    "https://example.com/beach1.jpg",
    "https://example.com/beach2.jpg"
  ],
  "duration": 8,
  "quality": "1080p",
  "aspect_ratio": "16:9",
  "generate_audio": false
}

Response Format

Generation Response

{
  "task_id": "task_abc123",
  "status": "pending",
  "created_at": "2024-01-01T12:00:00Z"
}

Status Response

{
  "task_id": "task_abc123",
  "status": "completed",
  "video_url": "https://cdn.example.com/video.mp4",
  "created_at": "2024-01-01T12:00:00Z",
  "completed_at": "2024-01-01T12:01:30Z"
}

Task Status Values

| Status | Description | Action Required | |--------|-------------|-----------------| | pending | Task queued | Continue polling | | processing | Generation in progress | Continue polling | | completed | Video ready | Retrieve video_url | | failed | Generation failed | Check error field |

Error Codes

HTTP Status Codes

| Code | Meaning | Common Causes | Solutions | |------|---------|---------------|-----------| | 200 | Success | - | Process response | | 400 | Bad Request | Invalid parameters, content blocked, file too large | Check parameters and content | | 401 | Unauthorized | Invalid or missing API key | Verify EVOLINK_API_KEY | | 402 | Payment Required | Insufficient balance | Add credits at dashboard | | 429 | Rate Limited | Too many requests | Wait and retry | | 503 | Service Unavailable | Server maintenance/overload | Retry later |

Error Response Format

{
  "error": "Detailed error message",
  "code": "ERROR_CODE",
  "details": {}
}

Common Error Messages

Content Blocking (400)

• Trigger: Realistic human faces, inappropriate content
• Message: Contains "face" or "人脸" keywords
• Solution: Modify prompt to avoid restricted content

File Size (400)

• Trigger: Images >30MB
• Message: Contains "file" and "large" or "size exceed"
• Solution: Compress images before upload

Invalid Key (401)

• Message: "Invalid API key" or similar
• Solution: Check key at https://evolink.ai/dashboard

Insufficient Balance (402)

• Message: "Insufficient balance" or similar
• Solution: Add credits at https://evolink.ai/dashboard

Polling Strategy

Recommended Pattern

1. Initial wait: Start polling immediately
2. Frequent polling: Every 5 seconds for first 30 seconds
3. Slower polling: Every 10 seconds after 30 seconds
4. Timeout: Stop after 3 minutes with warning

Typical Generation Times

• 4-5 seconds, 480p: 20-45 seconds
• 5-8 seconds, 720p: 30-90 seconds
• 10-15 seconds, 1080p: 60-180 seconds

Timeout Handling

After 3 minutes, inform user that generation may still be processing and suggest checking back later.

Rate Limits

• Generation requests: Varies by plan
• Status queries: Higher limit, safe to poll frequently
• Concurrent tasks: Varies by plan

Contact support for specific rate limit details for your account.

Output URLs

• Validity: 24 hours from generation
• Format: MP4 video files
• CDN delivery: High-speed download
• Audio: Synchronized audio included by default (voice, sound effects, background music)

Best Practices

Prompt Writing

• Be specific and descriptive
• Include visual details (colors, lighting, movement)
• Avoid realistic human faces
• Use cinematic language for better results

Performance Optimization

• Start with 480p for testing
• Use shorter durations for faster generation
• Provide clear, high-quality reference images
• Batch similar requests to optimize costs

Error Resilience

• Always handle all error codes
• Provide user-friendly error messages with action links
• Implement exponential backoff for rate limits
• Set reasonable timeouts for polling

Archive v1.4.0: 4 files, 9008 bytes

Files: _meta.json (139b), references/api-params.md (6220b), scripts/seedance-gen.sh (9056b), SKILL.md (5776b)

File v1.4.0:SKILL.md

---

name: seedance-2-video-gen description: Seedance 2.0 AI video generation via EvoLink API. Text-to-video, image-to-video with auto audio (voice, SFX, BGM). Works with OpenClaw, Claude Code, Cursor. Powered by ByteDance Seedance model. homepage: https://github.com/EvoLinkAI/evolink-skills metadata: {"openclaw":{"homepage":"https://github.com/EvoLinkAI/evolink-skills","requires":{"bins":["jq","curl"],"env":["EVOLINK_API_KEY"]},"primaryEnv":"EVOLINK_API_KEY"}}

Seedance Video Generation

An interactive AI video generation assistant powered by the Seedance model via EvoLink API.

After Installation

When this skill is first loaded, proactively greet the user and start the setup:

1. Check if EVOLINK_API_KEY is set

   • If not set: "To generate videos, you'll need an EvoLink API key. It takes 30 seconds to get one — just sign up at evolink.ai. Want me to walk you through it?"
   • If already set: "You're all set! What kind of video would you like to create?"

2. That's it. One question. The user is now in the flow.

Do NOT list features, show a menu, or dump instructions. Just ask one question to move forward.

Core Principles

1. Guide, don't decide — Present options and let the user decide. Don't make assumptions about their preferences.
2. Let the user drive the creative vision — If they have an idea, use their words. If they need inspiration, offer suggestions and let them choose or refine.
3. Smart context awareness — Recognize what the user has already provided and only ask about missing pieces.
4. Intent first — If the user's intent is unclear, confirm what they want before proceeding.

Flow

Step 1: Check for API Key

If the user hasn't provided an API key or set EVOLINK_API_KEY:

• Tell them they need an EvoLink API Key
• Guide them to register at https://evolink.ai and get a key from the dashboard
• Once they provide a key, proceed to Step 2

If the key is already set or provided, skip directly to Step 2.

Step 2: Understand Intent

Assess what the user wants based on their message:

• Intent is clear (e.g., "generate a video of a cat dancing") → Go to Step 3
• Intent is ambiguous (e.g., "I want to try Seedance") → Ask what they'd like to do: generate a new video, learn about model capabilities, etc.

Step 3: Gather Missing Information

Check what the user has already provided and only ask about what's missing:

| Parameter | What to tell the user | Required? | |-----------|----------------------|-----------| | Video content (prompt) | Ask what they'd like to see. If they need inspiration, suggest a few ideas for them to pick from or build on. | Yes | | Duration | Supported: 4–12 seconds. Ask how long they want. | Yes | | Resolution | Supported: 480p / 720p / 1080p. Ask their preference. | Yes | | Audio | The model can auto-generate voice, sound effects, and background music matching the video. Ask if they want audio enabled. | Yes | | Aspect ratio | Supported: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9. Only mention if relevant or if user asks. | Optional | | Reference images | Supported: up to 9 images (JPEG/PNG/WebP, ≤30MB each). Only mention if relevant. | Optional |

Smart gathering rules:

• User gives everything at once → Confirm and generate immediately
• User gives partial info → Only ask about the missing pieces
• User says "I want to generate a video" with no details → Guide from the beginning

Step 4: Generate

Once all required information is confirmed:

1. Tell the user: "Generating your video now — this usually takes 30–120 seconds. I'll let you know when it's ready."
2. Run the generation script. Do NOT forward each line of script output to the user. The script prints polling status internally — ignore it. Only report the final result.
3. When complete, share the video URL (valid for 24 hours) and generation time.

Script Usage

# Set API key
export EVOLINK_API_KEY=your_key_here

# Basic text-to-video
./scripts/seedance-gen.sh "user's prompt" --duration 5 --quality 720p

# With audio disabled
./scripts/seedance-gen.sh "user's prompt" --duration 8 --quality 1080p --no-audio

# With reference image
./scripts/seedance-gen.sh "user's prompt" --image "https://example.com/ref.jpg" --duration 6 --quality 720p

# Custom aspect ratio
./scripts/seedance-gen.sh "user's prompt" --aspect-ratio 9:16 --duration 4 --quality 480p

Error Handling

Provide friendly, actionable messages:

| Error | What to tell the user | |-------|----------------------| | Invalid/missing key (401) | "Your API key doesn't seem to work. You can check it at https://evolink.ai/dashboard" | | Insufficient balance (402) | "Your account balance is low. You can add credits at https://evolink.ai/dashboard" | | Rate limited (429) | "Too many requests — let's wait a moment and try again" | | Content blocked (400) | "This prompt was flagged (realistic human faces are restricted). Try adjusting the description" | | Service unavailable (503) | "The service is temporarily busy. Let's try again in a minute" |

Model Capabilities Summary

Use this when the user asks what the model can do:

• Text-to-video: Describe a scene, get a video
• Image-to-video: Provide reference images to guide the output
• Audio generation: Auto-generates synchronized voice, sound effects, and background music
• Duration: 4–12 seconds
• Resolution: 480p, 720p, 1080p
• Aspect ratios: 16:9, 9:16, 1:1, 4:3, 3:4, 21:9
• Limitation: Realistic human faces are restricted

References

• references/api-params.md: Complete API parameter reference
• scripts/seedance-gen.sh: Generation script with automatic polling and error handling

File v1.4.0:_meta.json

{ "ownerId": "kn74p4xy6sja0199cea53anecs81kqjs", "slug": "seedance-2-video-gen", "version": "1.4.0", "publishedAt": 1771868583552 }

File v1.4.0:references/api-params.md

Seedance API Parameters Reference

This document provides complete API parameter reference for the Seedance video generation service.

API Endpoints

Generation Request

POST https://api.evolink.ai/v1/videos/generations
Authorization: Bearer {EVOLINK_API_KEY}
Content-Type: application/json

Task Status Query

GET https://api.evolink.ai/v1/tasks/{task_id}
Authorization: Bearer {EVOLINK_API_KEY}

Generation Parameters

Required Parameters

| Parameter | Type | Description | |-----------|------|-------------| | model | string | Fixed: "seedance-1.5-pro" | | prompt | string | Text description of desired video content |

Optional Parameters

| Parameter | Type | Default | Options | Description | |-----------|------|---------|---------|-------------| | image_urls | array | [] | URLs | Reference images (max 9) | | duration | integer | 5 | 4-12 | Video duration in seconds | | quality | string | "720p" | "480p", "720p", "1080p" | Video resolution | | aspect_ratio | string | "16:9" | "16:9", "9:16", "1:1", "4:3", "3:4", "21:9" | Video aspect ratio | | generate_audio | boolean | true | true, false | Generate synchronized audio (voice, SFX, music). Enabled by default; disabling reduces cost. |

Parameter Constraints

Image URLs

• Maximum count: 9 images
• File size limit: ≤30MB per image
• Supported formats: JPEG, PNG, WebP, BMP, TIFF, GIF
• Format: Array of strings containing full URLs

Duration

• Range: 4-12 seconds
• Type: Integer
• Impact: Longer videos cost more and take longer to generate

Quality

• 480p: Fastest, cheapest, lower quality
• 720p: Balanced (recommended default)
• 1080p: Highest quality, slower, most expensive

Aspect Ratio

• 16:9: Standard landscape (YouTube, TV)
• 9:16: Mobile/vertical (TikTok, Instagram Stories)
• 1:1: Square (Instagram posts, social media)
• 4:3: Traditional format
• 3:4: Portrait format
• 21:9: Ultra-wide cinematic

Example Payloads

Text-to-Video

{
  "model": "seedance-1.5-pro",
  "prompt": "A serene sunset over calm ocean waves",
  "duration": 5,
  "quality": "720p",
  "aspect_ratio": "16:9",
  "generate_audio": false
}

Image-to-Video

{
  "model": "seedance-1.5-pro", 
  "prompt": "Gentle waves rolling onto the shore",
  "image_urls": [
    "https://example.com/beach1.jpg",
    "https://example.com/beach2.jpg"
  ],
  "duration": 8,
  "quality": "1080p",
  "aspect_ratio": "16:9",
  "generate_audio": false
}

Response Format

Generation Response

{
  "task_id": "task_abc123",
  "status": "pending",
  "created_at": "2024-01-01T12:00:00Z"
}

Status Response

{
  "task_id": "task_abc123",
  "status": "completed",
  "video_url": "https://cdn.example.com/video.mp4",
  "created_at": "2024-01-01T12:00:00Z",
  "completed_at": "2024-01-01T12:01:30Z"
}

Task Status Values

| Status | Description | Action Required | |--------|-------------|-----------------| | pending | Task queued | Continue polling | | processing | Generation in progress | Continue polling | | completed | Video ready | Retrieve video_url | | failed | Generation failed | Check error field |

Error Codes

HTTP Status Codes

| Code | Meaning | Common Causes | Solutions | |------|---------|---------------|-----------| | 200 | Success | - | Process response | | 400 | Bad Request | Invalid parameters, content blocked, file too large | Check parameters and content | | 401 | Unauthorized | Invalid or missing API key | Verify EVOLINK_API_KEY | | 402 | Payment Required | Insufficient balance | Add credits at dashboard | | 429 | Rate Limited | Too many requests | Wait and retry | | 503 | Service Unavailable | Server maintenance/overload | Retry later |

Error Response Format

{
  "error": "Detailed error message",
  "code": "ERROR_CODE",
  "details": {}
}

Common Error Messages

Content Blocking (400)

• Trigger: Realistic human faces, inappropriate content
• Message: Contains "face" or "人脸" keywords
• Solution: Modify prompt to avoid restricted content

File Size (400)

• Trigger: Images >30MB
• Message: Contains "file" and "large" or "size exceed"
• Solution: Compress images before upload

Invalid Key (401)

• Message: "Invalid API key" or similar
• Solution: Check key at https://evolink.ai/dashboard

Insufficient Balance (402)

• Message: "Insufficient balance" or similar
• Solution: Add credits at https://evolink.ai/dashboard

Polling Strategy

Recommended Pattern

1. Initial wait: Start polling immediately
2. Frequent polling: Every 5 seconds for first 30 seconds
3. Slower polling: Every 10 seconds after 30 seconds
4. Timeout: Stop after 3 minutes with warning

Typical Generation Times

• 4-5 seconds, 480p: 20-45 seconds
• 5-8 seconds, 720p: 30-90 seconds
• 10-15 seconds, 1080p: 60-180 seconds

Timeout Handling

After 3 minutes, inform user that generation may still be processing and suggest checking back later.

Rate Limits

• Generation requests: Varies by plan
• Status queries: Higher limit, safe to poll frequently
• Concurrent tasks: Varies by plan

Contact support for specific rate limit details for your account.

Output URLs

• Validity: 24 hours from generation
• Format: MP4 video files
• CDN delivery: High-speed download
• Audio: Synchronized audio included by default (voice, sound effects, background music)

Best Practices

Prompt Writing

• Be specific and descriptive
• Include visual details (colors, lighting, movement)
• Avoid realistic human faces
• Use cinematic language for better results

Performance Optimization

• Start with 480p for testing
• Use shorter durations for faster generation
• Provide clear, high-quality reference images
• Batch similar requests to optimize costs

Error Resilience

• Always handle all error codes
• Provide user-friendly error messages with action links
• Implement exponential backoff for rate limits
• Set reasonable timeouts for polling