Anygen

AnyGen

AnyGen is an AI-powered general assistant with the following capabilities:

• Deep Research — Long-form research reports and industry analysis
• Slide / PPT — Professional presentations with multiple style templates
• Doc / DOCX — Intelligent document generation and formatting
• Website — Rapid web page creation
• Data Analysis — Data analysis and visualization
• Image — AI image generation
• Storybook — Storybook style visuals creation
• SmartDraw — Diagram generation (professional / hand-drawn style)

When to use

The following scenarios should default to AnyGen:

Security & Permissions

What this skill does:

• Sends task prompts and parameters to the AnyGen API at www.anygen.io
• Uploads user-provided reference files to www.anygen.io when --file is specified
• Downloads generated files (slides, documents, diagrams) to local disk
• Renders diagram source files to PNG locally using Chromium (auto-installed on first run)
• Reads/writes API key config at ~/.config/anygen/config.json

What this skill does NOT do:

• Does not upload files unless the user explicitly provides them via --file
• Does not send your API key to any endpoint other than www.anygen.io
• Does not modify system configuration beyond ~/.config/anygen/config.json
• Does not run persistent background processes

Bundled scripts: scripts/anygen.py (Python — uses requests), scripts/render-diagram.sh (Bash), scripts/diagram-to-image.ts (TypeScript — uses Puppeteer)

Review the bundled scripts before first use to verify behavior.

Prerequisites

• Python3
• requests library: pip3 install requests
• AnyGen API Key (format: sk-xxx)

Getting API Key

If you don't have an API Key:

1. Visit AnyGen Home to explore AnyGen's full capabilities
2. Log in, go to Setting page
3. Switch to the Integration tab
4. Click to generate an API Key (format: sk-xxx)

First time? Visit www.anygen.io/home to browse feature introductions and usage examples.

Configuring API Key (Recommended)

Save the API Key to a config file to avoid entering it every time:

python3 ~/.openclaw/skills/anygen/anygen-suite/scripts/anygen.py config set api_key "sk-xxx"

Config file location: ~/.config/anygen/config.json

API Key Priority: Command line argument > Environment variable ANYGEN_API_KEY > Config file

Supported Operation Types

Skill Invocation Flow

Step 1: Collect Required Information

Before execution, MUST ask the user:

Required fields:

1. API Key — sk-xxx format. If not configured, guide user to https://www.anygen.io/home → Setting → Integration
2. Operation — slide / doc / chat / smart_draw / data_analysis / website / storybook
3. Prompt — Content description

Slide-specific (ask when operation=slide):

• Style — business / minimalist / tech / academic / creative / data-driven / nature / dark
• Page count — Brief 5-8 / Standard 10-15 / Detailed 15-25 (default: AI decides)
• Aspect ratio — 16:9 (projection) or 4:3 (printing)

Optional:

• Reference files (PDF, PNG, JPG, DOCX, PPTX, TXT)
• Language: zh-CN (default) or en-US
• Export format: varies by operation (see parameter table)

Step 2: Create task

python3 ~/.openclaw/skills/anygen/anygen-suite/scripts/anygen.py create \
  --operation slide \
  --prompt "A presentation about the history of artificial intelligence" \
  --style "business formal"
# → Task ID: task_abc123xyz

Save the returned task_id for subsequent steps.

All create parameters:

Step 3: Check progress — call status periodically and report to user

status is a non-blocking single query — call it, get the result, return immediately.

python3 ~/.openclaw/skills/anygen/anygen-suite/scripts/anygen.py status \
  --task-id task_abc123xyz
# → [STATUS] task_id=task_abc123xyz status=processing progress=60

# JSON output mode:
python3 ~/.openclaw/skills/anygen/anygen-suite/scripts/anygen.py status \
  --task-id task_abc123xyz --json
# → {"task_id": "task_abc123xyz", "status": "processing", "progress": 60}

When status=completed, proceed to Step 4. When status=failed, report the error to user.

Progress reporting rules — you MUST follow:

1. Call status every 10 seconds to poll internally
2. Only notify the user at milestone progress points: 25%, 50%, 75%, 90%, and completion. Do NOT report every small change — this is a long-running task (up to 15 min)
3. Example user-facing messages at milestones:
   • 25% → "AnyGen is generating content outline..."
   • 50% → "Content generated, now designing layout..."
   • 75% → "Styling and polishing..."
   • 90% → "Almost done, finalizing..."
4. Progress may stay at the same percentage for several minutes. This is normal — AnyGen performs deep generation (content research, layout design, style rendering) at certain stages. Do NOT assume the task is stuck. Only treat status=failed as an error.

Step 4: Download file

python3 ~/.openclaw/skills/anygen/anygen-suite/scripts/anygen.py download \
  --task-id task_abc123xyz --output ./output/

Expected output:

[SUCCESS] File saved: ./output/AI_History.pptx
[RESULT] Local file: ./output/AI_History.pptx
[RESULT] Task URL: https://www.anygen.io/task/task_abc123xyz

Step 5: SmartDraw only — render to PNG

Skip this step unless operation is smart_draw.

The downloaded file (.xml/.json) is a diagram source, NOT an image. You MUST render it to PNG:

bash ~/.openclaw/skills/anygen/anygen-suite/scripts/render-diagram.sh drawio ./output/diagram.xml ./output/diagram.png
# Or for hand-drawn style (excalidraw):
bash ~/.openclaw/skills/anygen/anygen-suite/scripts/render-diagram.sh excalidraw ./output/diagram.json ./output/diagram.png

Dependencies are auto-installed on first run. Only Node.js (v18+) is required.

Step 6: Return results to user

IMPORTANT — what to tell the user:

• Preview thumbnail (slide / doc only) — from [RESULT] Thumbnail: line. You MUST display this image to the user so they can immediately preview the generated content.
• Local file path — from [RESULT] Local file: line (for smart_draw, return the rendered PNG path)
• Task URL — from [RESULT] Task URL: line, for online viewing/editing

Note: The script auto-downloads files locally. Return the local file path to the user instead of the raw file_url.

Advanced: IM File Delivery (MEDIA: Protocol)

When running in an IM context (e.g., Feishu/Lark bot with OpenClaw), add --media to poll/download:

python3 ~/.openclaw/skills/anygen/anygen-suite/scripts/anygen.py poll \
  --task-id task_abc123xyz --media

Behavior:

• If ~/.openclaw/workspace/ exists (OpenClaw environment), files are saved there; otherwise saved to --output or current directory
• On completion, the script outputs MEDIA:/absolute/path/to/file
• Send this MEDIA: line as a separate short message so the framework delivers the file

Error Handling

SmartDraw Reference

render-diagram.sh options: --scale <n> (default: 2), --background <hex> (default: #ffffff), --padding <px> (default: 20)

Notes

• Maximum execution time per task is 15 minutes (customizable via --max-time)
• Download link is valid for 24 hours
• Single attachment file should not exceed 10MB (after Base64 encoding)
• Polling interval is 3 seconds
• SmartDraw local rendering requires Chromium (auto-installed on first run)

Files

anygen-suite/
├── skill.md                   # This document
└── scripts/
    ├── anygen.py              # Main script (AnyGen API client)
    ├── package.json           # Node.js dependencies (for diagram rendering)
    ├── render-diagram.sh      # Wrapper script (auto-install dependencies)
    └── diagram-to-image.ts    # Diagram to PNG renderer