alibabacloud-yike-storyboard
Yike Storyboard Creation Skill
Help users complete the AI video creation workflow from novel/script to storyboard via conversational interface.
Prerequisites
1. Aliyun CLI Check
Pre-check: Aliyun CLI >= 3.3.1 required
aliyun versionVerify version >= 3.3.1. If not installed or version is too low, refer to references/cli-installation-guide.md.
Then [MUST] run the following commands to configure CLI:
# Enable auto plugin installation aliyun configure set --auto-plugin-install true # Set region to cn-shanghai (required for ICE service) aliyun configure set --region cn-shanghai
2. ICE Plugin Installation
Pre-check: ICE CLI Plugin required
aliyun plugin install --names iceVerify plugin is installed:
aliyun ice --help
Service Activation
1. Activate IMS (Intelligent Media Services)
Activation Link: https://ice.console.aliyun.com/guide/default
Log in with your Alibaba Cloud account and activate ICE Intelligent Media Services.
2. Log in to yikeai platform
Platform Link: https://www.yikeai.com/#/home
Visit the product website and authorize login with your Alibaba Cloud account. Currently Wanjing Yike is in invitation-only phase. Click Apply for Access to submit your application.
Credential Verification
Pre-check: Alibaba Cloud Credentials Required
Security Rules:
- DO NOT read, print, or output AK/SK values
- ONLY USE
aliyun configure listto check credential statusaliyun configure listIf no valid profile exists, STOP and:
- Get credentials from Alibaba Cloud Console
- Run
aliyun configureto set up credentials
RAM Permissions
This skill requires ICE permissions: ice:CreateYikeAssetUpload, ice:SubmitYikeStoryboardJob, ice:GetYikeStoryboardJob.
For complete permission policies, see references/ram-policies.md.
[MUST] Permission Error Handling: When any command fails due to permission errors, read
references/ram-policies.mdfor required permissions.
Parameter Confirmation
Confirm key parameters with user before file upload through natural dialogue.
| Parameter | Required | Description | Default |
|---|---|---|---|
| file_path | ✅ | Text file path (txt/docx, ≤5MB, ≤30K chars) | User provides |
| title | ✅ | Storyboard title | Extract from text |
| source-type | ✅ | Novel or Script |
Based on content analysis |
| style | ✅ | Visual style ID | Based on genre |
| voice | ✅ | Narration voice ID | Based on protagonist |
| shot-split-mode | ✅ | Shot split mode | Based on narrative style |
| ratio | Optional | 16:9, 9:16, 4:3, 3:4 |
9:16 |
| resolution | Optional | 720P, 1K, 2K, 4K |
720P |
See Task 0: Parameter Confirmation for recommendation guide.
Text Type Classification
| Type | Value | Description |
|---|---|---|
| Novel | Novel |
Primarily narrative, descriptive, psychological content for reading |
| Script | Script |
Primarily scenes, dialogue, action descriptions for performance/filming |
Classification Guide:
Evaluate the following features to determine if text is novel or script (don't judge solely by dialogue presence - novels can have extensive dialogue too):
| Feature | Script | Novel |
|---|---|---|
| Scene Markers | ✅ Has scene numbers, time, location, INT/EXT | ❌ No explicit scene markers |
| Structure | ✅ "Character Name + Dialogue" dominant | ❌ Narrative text dominant |
| Action Cues | ✅ Has stage directions, camera directions | ❌ No performable features |
| Literary Expression | ❌ Minimal | ✅ Rich environmental descriptions, psychology, emotions |
Classification Rules:
- Script: Dominated by scenes, dialogue, action descriptions
- Novel: Dominated by narration, description, psychological activity
- Mixed Content: Determine dominant feature, output "more like script" or "more like novel"
Important: After classification, confirm text type with user or let user specify directly.
Core Workflow
Task 0: Parameter Confirmation
Before upload, analyze the text and confirm key parameters with the user.
Step 1: Analyze Text Content
head -c 1000 <file_path>
Determine: genre, narrative style (first/third person), protagonist characteristics.
Step 2: Recommend & Confirm Parameters
Based on analysis, make recommendations and confirm with user through natural dialogue.
Recommendation Guide:
| Parameter | Based On | Examples |
|---|---|---|
| style | Genre | Modern Urban → CinematicRealism; Period Drama → RealisticGuzhuangPro; Fantasy → RealisticXianxia; Anime → Ghibli |
| voice | Protagonist | Young Female → sys_ClassicYoungWoman; Young Male → sys_GentleYoungMan; Mature Male → sys_CalmDeepMale |
| shot-split-mode | Narrative | Third person → thirdPersonNarration; First person → firstPersonNarration; Dialogue-heavy Script → dialogue |
| ratio | Platform | TikTok/Douyin → 9:16; YouTube → 16:9 |
| resolution | Quality | Default 720P; Higher quality 1K/2K/4K |
Defaults: 9:16, 720P (vertical HD for mobile)
⚠️ Constraint:
dialoguemode is ONLY available forScripttype, NOT forNovel.
Example confirmation dialogue:
"Based on your modern urban romance novel, I recommend:
- Style:
CinematicRealism(film-quality, great for emotional scenes)- Voice:
sys_ClassicYoungWoman(matches your young female protagonist)- Shot Mode:
thirdPersonNarration(for third-person narrative)- Format:
9:16,720P(vertical HD for mobile platforms)- Title: "雨夜归途"
Does this look good, or would you like to change anything?"
If user has no preference, use recommended defaults. For full options, see Style Mapping Table and Voice Mapping Table.
DO NOT proceed to Task 1 until user confirms.
Task 1: Upload Text File to OSS
Prerequisite: Task 0 (Parameter Confirmation) MUST be completed.
Use helper script to automatically get credentials and upload:
bash scripts/upload_to_oss.sh <file_path>
Returns: FileURL (for subsequent job submission)
Task 2: Submit Storyboard Job
aliyun ice submit-yike-storyboard-job \
--file-url "<FileURL>" \
--source-type <SourceType> \
--style-id <StyleId> \
--narration-voice-id <VoiceId> \
--aspect-ratio "9:16" \
--resolution 720P \
--shot-split-mode <ShotSplitMode> \
--shot-prompt-mode multi \
--video-model "wan2.6-r2v-flash" \
--exec-mode StoryboardOnly \
--title "<Title>" \
--region cn-shanghai \
--user-agent AlibabaCloud-Agent-Skills
For complete parameter reference, see references/related-commands.md.
Task 3: Query Job Status
aliyun ice get-yike-storyboard-job \
--job-id <JobId> \
--region cn-shanghai \
--user-agent AlibabaCloud-Agent-Skills
For complete parameter and response reference, see references/related-commands.md.
Status Flow and User Prompts:
| Status | SubStatus | Current Phase | User Prompt | Action |
|---|---|---|---|---|
| Configuring | Parsing | Entity Asset Parsing | "Parsing your script, AI is extracting characters, scenes and props..." | Wait |
| Configuring | ParseSucc | Entity Asset Image Generation | "Script parsed! Now generating images for characters, scenes and props. You can preview the progress here:" | Provide Entity Management Link |
| Editing | Creating | Shot Script Generation | "Entity assets ready! Now creating shot script, almost done..." | Wait |
| Editing | CreateSucc | Complete ✅ | "Shot script complete! You can now view and edit your storyboard:" | Provide Editing Link |
| Editing | CreateFailed | Failed ❌ | "Shot script generation failed, please check error message or resubmit." | Troubleshoot |
Interactive Links:
| Status | Link | Purpose |
|---|---|---|
| ParseSucc | https://www.yikeai.com/#/storyboard/entitiesManagement?storyboardId={storyboardId} |
Preview entity assets (characters, scenes, props) generation progress |
| CreateSucc | https://www.yikeai.com/#/storyboard/editing?storyboardId={storyboardId} |
Edit storyboard, generate videos, export final video |
Job Status Description:
| JobStatus | Description | User Prompt |
|---|---|---|
| Running | Job in progress | "Job is processing, usually takes a few minutes, please wait..." |
| Succeeded | Job succeeded | "Job completed!" |
| Failed | Job failed | "Job failed, please check error message." |
| Suspended | Job suspended | "Job suspended, some shots failed, can be manually fixed in storyboard." |
Query Recommendation: Jobs usually take a few minutes. Recommend querying status every 30 seconds.
Task 4: Get Storyboard Link
Get storyboardId from JobResult.StoryboardInfoList and construct link:
https://www.yikeai.com/#/storyboard/editing?storyboardId={storyboardId}
Style Mapping Table
| StyleId | Name |
|---|---|
| RealisticPhotographyPro | Realistic Photography Pro |
| RealisticGuzhuangPro | Realistic Chinese Period Pro |
| RealisticPhotography | Realistic Photography |
| RealisticGuzhuang | Realistic Chinese Period |
| RealisticXianxia | Realistic Xianxia |
| RealisticEra | Realistic Period |
| RealisticWasteland | Realistic Wasteland |
| GuofengAnime | 2D Chinese Style Anime |
| GuofengAnime3D | 3D Chinese Style Anime |
| Cartoon3D | 3D Cartoon |
| Photorealistic3D | Photorealistic 3D Render |
| SciFiRealism | Sci-Fi Realism |
| Chibi3D | 3D Chibi |
| ShojoManga | Shojo Manga |
| NewPeriodAnime | New Era Anime |
| FairyTale2D | 2D Fairy Tale |
| Wasteland2D | 2D Wasteland |
| InkWuxia | Ink Wuxia |
| ShadiaoMeme | Panda Meme Style |
| Chibi2D | 2D Chibi |
| Ghibli | Ghibli |
| SciFiComic | Cyberpunk |
| AmericanSuperhero | American Superhero |
| Hokusei | Hokusei |
| RealisticComic | Realistic Comic |
| CinematicRealism | Cinematic Realism |
| MinimalistRealism | Minimalist Realism |
| ShonenManga | Shonen Manga |
Narration Voice Mapping Table
| Voice ID | Description |
|---|---|
| sys_ClassicMiddleAgedWoman | Classic Female Narrator (25-45, wise) |
| sys_ClassicYoungWoman | Classic Young Female (18-25, intellectual) |
| sys_IntellectualYoungWoman | Intellectual Young Female (18-25, intellectual) |
| sys_GentleYoungMan | Gentle Young Male (18-25, gentle) |
| sys_WiseYoungMan | Wise Young Male (18-25, wise) |
| sys_ClassicYoungMan | Classic Young Male (18-25, charming) |
| sys_thoughtfulBoy | Thoughtful Boy (10-15, well-behaved) |
| sys_SereneIntellect | Serene Intellectual Male (18-25, cool and rational) |
| sys_RichBassMale | Rich Bass Male (18-25, deep voice) |
| sys_CalmDeepMale | Calm Deep Male (25-40, steady and deep) |
| sys_MajesticBaritone | Majestic Baritone (40-60, authoritative) |
| sys_GravellySoulful | Gravelly Soulful Male (40-60, weathered) |
| sys_SweetBrightGirl | Sweet Bright Girl (10-15, lively) |
| sys_GracefulPoisedWoman | Graceful Poised Woman (18-25, elegant) |
| longbaizhi | Long Baizhi (20-30, witty female narrator) |
| sys_YoungGracefulWoman | Young Graceful Woman (18-25, gentle) |
| sys_MaturePoisedWoman | Mature Poised Woman (25-40, graceful) |
| sys_MatureWiseWoman | Mature Wise Woman (25-40, elegant and wise) |
| sys_ElderlyWistfulWoman | Elderly Wistful Woman (40-60, nostalgic) |
Shot Split Mode
| Mode | Description | Use Case | Supported Type |
|---|---|---|---|
| dialogue | Dialogue Mode | Dialogue-heavy scripts, short dramas | Script only |
| firstPersonNarration | First Person Narration | Stories from protagonist's perspective, diary style | Novel/Script |
| firstPersonNarrationPureVO | First Person Pure VO | Inner monologue stories, prose | Novel/Script |
| thirdPersonNarration | Third Person Narration | Omniscient perspective stories, fairy tales, historical | Novel/Script |
Restriction: When
source-typeisNovel,dialoguemode is NOT supported.
Recommendation Guide:
Choose appropriate mode based on text type and content analysis:
Script:
- High dialogue ratio (short dramas, chat format, comedy) → Recommend
dialogue - First person narration → Recommend
firstPersonNarration - Pure inner monologue/narration (prose, reflections) → Recommend
firstPersonNarrationPureVO - Third person narration → Recommend
thirdPersonNarration
Novel:
- "I" perspective + has dialogue (urban romance, mystery) → Recommend
firstPersonNarration - Pure inner monologue/narration (prose, reflections) → Recommend
firstPersonNarrationPureVO - Third person narration (fairy tales, mythology, history) → Recommend
thirdPersonNarration
Important: Before execution, describe mode features to user and let them confirm or choose.
Capability Scope
This skill automates:
- Upload novel/script files to OSS
- Entity asset parsing (characters, scenes, props extraction)
- Entity asset image generation
- Shot script generation
- Job status tracking
After shot script completion, continue in UI:
Once the job reaches CreateSucc status, this skill's automation is complete. The following steps require the Yike Storyboard web interface:
- Edit shot script - Adjust shot content, descriptions, and prompts
- Generate shot images/videos - AI generates images/videos for each shot
- Edit and assemble - Fine-tune timing, add transitions
- Export final video - Render and download the complete video
Next Step: Open the storyboard editing link and continue your video creation journey!
Reference Links
| Reference | Description |
|---|---|
| references/cli-installation-guide.md | CLI Installation Guide |
| references/ram-policies.md | RAM Permission Policies |
| references/related-commands.md | Related CLI Commands |
| references/verification-method.md | Verification Methods |
Error Handling
| Error | Cause | Solution |
|---|---|---|
| MainAccountUserNotFound | Yike service not activated | Apply for access at Service Activation section |
| InvalidAccessKeyId | Invalid AK/SK | Check credential configuration |
| Forbidden | Insufficient permissions | See RAM Permissions section |
| region can't be empty | OSS upload missing region | Add --region cn-shanghai |
Note: If you receive
MainAccountUserNotFounderror, it means your account has not been whitelisted for the Yike service. Please visit https://www.yikeai.com and apply for access through the invitation application form.