nocobase-ai-employee

Guide AI to create and manage NocoBase AI employees — chatbot assistants with page integration

Safety Notice

This listing is from the official public ClawHub registry. Review SKILL.md and referenced scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "nocobase-ai-employee" with this command: npx skills add alexander-lq/aicreat

NocoBase AI Employee Management

You are guiding the user to create and manage AI employees (chatbot assistants) in NocoBase.

Key Concepts

AI Employee

An AI employee is a chatbot assistant configured with:

  • username (PK): Unique identifier, e.g. am-asset-keeper
  • nickname: Display name shown to users
  • about: System prompt defining role, data scope, and behavior
  • skills: Tool bindings (query, count, form fill, workflow call)
  • modelSettings: LLM configuration (service, model, temperature)

Page Integration

AI employees appear on pages in two ways:

  1. Floating Avatar (AIEmployeeShortcutListModel + AIEmployeeShortcutModel): Circular avatar buttons in page top-right corner
  2. Action Bar Button (AIEmployeeButtonModel): AI button in table/form action bars

Workflow

Phase 1: Create AI Employees

nb_create_ai_employee("my-helper", "助手", "通用助手",
    "nocobase-015-male", "One-line description",
    "Full system prompt with role, data scope, behavior rules...",
    "Welcome message...",
    '[{"name":"dataSource-dataSourceQuery","autoCall":true}]')

Phase 2: Add Page Shortcuts (Floating Avatars)

nb_ai_shortcut("tab_uid", '[
    {"username": "my-helper", "tasks": [
        {"title": "Quick Query", "message": {"user": "帮我查询数据"}, "autoSend": false}
    ]}
]')

Phase 3: Add Block Buttons

nb_ai_button("table_uid", "my-helper", '[
    {"title": "Analyze Data", "message": {"user": "分析当前数据"}, "autoSend": false}
]')

Available Skills (Tool Bindings)

Tool NameDescriptionautoCall
dataModeling-getCollectionNamesDiscover table namestrue
dataModeling-getCollectionMetadataGet field definitionstrue
dataSource-dataSourceQueryQuery databasetrue
dataSource-dataSourceCountingCount recordstrue
frontend-formFillerAuto-fill formstrue
workflowCaller-<key>Custom workflow toolfalse

Avatar IDs

Common avatar IDs: nocobase-001-male through nocobase-060-male, nocobase-001-female through nocobase-060-female.

Model Settings

Default LLM configuration:

{
  "llmService": "gemini",
  "model": "models/gemini-2.5-flash",
  "temperature": 0.7,
  "topP": 1,
  "timeout": 60000,
  "maxRetries": 1,
  "responseFormat": "text"
}

Common adjustments:

  • temperature: 0.3 — more deterministic, better for data queries
  • temperature: 0.9 — more creative, better for writing/suggestions
  • timeout: 120000 — longer timeout for complex multi-step queries
  • responseFormat: "markdown" — for formatted output

Page Integration — Complete Example

Floating Avatar (Shortcut)

Place AI employee avatars on a page for quick access:

nb_ai_shortcut("tab_uid", '[
    {"username": "my-helper", "tasks": [
        {"title": "Quick Query", "message": {"user": "帮我查一下最新数据"}, "autoSend": false},
        {"title": "Generate Report", "message": {"user": "生成本月汇总报表"}, "autoSend": false}
    ]}
]')

Multiple employees on the same page:

nb_ai_shortcut("tab_uid", '[
    {"username": "data-analyst", "tasks": [...]},
    {"username": "form-helper", "tasks": [...]}
]')

Block Button

Add AI action button to a table's action bar:

nb_ai_button("table_uid", "my-helper", '[
    {"title": "Analyze Selected", "message": {"user": "分析当前选中的记录"}, "autoSend": false}
]')

Best Practices

  1. Role-focused: Each AI employee covers one business domain
  2. Progressive: Start with basic query tools, add workflow tools later
  3. Chinese-first: Use {{$nLang}} in system prompts for language awareness
  4. Data scope: Explicitly list accessible tables in the system prompt
  5. Preset tasks: Add 1-2 quick-start tasks to shortcuts and buttons
  6. System prompt structure: Role → Data scope (table names) → Behavior rules → Output format
  7. Skill selection: Start with dataSource-dataSourceQuery + dataModeling-getCollectionMetadata for data-oriented employees

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

Automation

Discord

Use when you need to control Discord from Clawdbot via the discord tool: send messages, react, post or upload stickers, upload emojis, run polls, manage threads/pins/search, fetch permissions or member/role/channel info, or handle moderation actions in Discord DMs or channels.

Registry SourceRecently Updated
33.6K72steipete
Automation

AgentCall

Give your agent real phone numbers for SMS, OTP verification, and voice calls via the AgentCall API.

Registry SourceRecently Updated
Automation

clawbus-skill

clawbus skill marketplace for AI agents. Search, download, install, and activate skills from the clawbus library. Use when the user asks for a capability you...

Registry SourceRecently Updated
Automation

chat2workflow

A design-only workflow designer for the Dify and Coze platforms. Through multi-round conversation, it produces a structured workflow JSON (nodes, edges, vari...

Registry SourceRecently Updated