nocobase-ai-employee
Use when a NocoBase task requires AI employee lifecycle work such as discovering existing employees, judging fit, creating a dedicated employee, or configuring profile, prompt, model, skills, tools, or knowledge base before another skill binds it to a UI surface.
What this skill does
# NocoBase AI Employee
## Goal
Configure an AI employee-backed action for a specific NocoBase business surface. Decide whether the requirement should use built-in UI actions, JS actions, workflows, or an AI employee; then reuse or create the right employee and hand off the final UI placement to `nocobase-ui-builder`.
## Required Hand-Off Skills
- Use `nocobase-ui-builder` for Modern page/block/action authoring and AI employee action placement.
- Use `nocobase-data-modeling` when the requirement needs new collections, fields, or relations before the AI action can work.
- Use `nocobase-workflow-manage` when the AI employee should call or trigger a workflow tool, or when the task is mostly deterministic backend automation.
## Decision Gate
Classify the user request before writing anything:
| Need | Prefer |
| --- | --- |
| Fixed CRUD, navigation, visibility, filters, field assignment, or simple button behavior | Built-in UI action / reaction through `nocobase-ui-builder` |
| Deterministic client-side calculation, formatting, validation, or data transform | JS action / JS surface through `nocobase-ui-builder` |
| Deterministic multi-step server automation, approval, notification, scheduled work | Workflow through `nocobase-workflow-manage` |
| Natural-language interpretation, ambiguous intent, extraction from messy text, summarization, classification, drafting, recommendations, data insight narrative, tool choice, or model judgment | AI employee action |
Only choose AI employee when model judgment materially reduces ambiguity or gives the user a natural-language task surface. Do not use AI employee as a substitute for ordinary deterministic UI configuration.
## Workflow
1. **Decompose the request**
- Identify the target page/block/action slot, target collection, current-record vs whole-block context, and expected user interaction.
- Split deterministic setup from non-deterministic model work.
- If the target UI surface is not uniquely known, use `nocobase-ui-builder` inspection routes first.
2. **Decide if AI employee is appropriate**
- Apply the Decision Gate.
- If AI is not appropriate, hand off to the relevant skill and explain the narrower route.
- If AI is appropriate, write a compact task contract: employee role, input context, expected output, whether to auto-send, whether to use web search, and any required tools/skills.
3. **Discover existing AI employees**
- Prefer user-visible employees with `aiEmployees:listByUser`.
- If using admin list data, filter to `enabled=true`, `deprecated=false`, and role-visible rows. Avoid `category="developer"` unless the user is configuring developer-facing builder work.
- Match by role, position, bio, existing tools, and `modelSettings`.
- Read `references/ai-employee-api.md` only when you need concrete resource names, fields, or payload shapes.
4. **Reuse or create**
- Reuse an existing employee when one clearly covers the role with compatible tools and model restrictions.
- Create a new employee only when no existing employee reaches roughly 70% fit, or when the user explicitly wants a dedicated employee.
- For new employees, keep `bio` human-facing and put operational behavior in `about`.
- For new employees, validate the create payload has `avatar` set to a supported avatar seed before calling `aiEmployees:create`.
- If `avatar` is missing, empty, null, or unsupported, set it to the default supported seed `nocobase-015-male`.
- Do not create developer-category employees for business users unless explicitly requested.
5. **Bind the employee to the block**
- Use `nocobase-ui-builder` and its AI employee action reference.
- Use public action shape only: `type: "aiEmployee"` with `settings.username`, `settings.auto`, `settings.workContext`, `settings.tasks`, `settings.style`.
- Do not write raw `props`, `stepParams`, `flowModels`, or database rows.
- For block/form/record context, default to `workContext: [{ "type": "flow-model", "target": "self" }]`.
6. **Verify**
- Read back or inspect the target surface through `nocobase-ui-builder` when a write occurred.
- Verify the AI action points at the intended username and has the intended task message.
- If a new employee was created, verify it appears in `aiEmployees:listByUser` for the intended role, and verify its `avatar` is non-empty and still one of the supported avatar seeds. Explain any role/ACL follow-up if it is not visible.
## Employee Matching Rules
Prefer built-ins when they fit:
- `atlas`: route a broad request to other employees or coordinate sub-agents.
- `dex`: extract, clean, structure, or fill forms from messy text.
- `viz`: analyze data and produce insights or reports.
- `ellis`: understand, summarize, and draft email replies.
- `lexi` / `lina`: translation and localization, with `lina` usually developer-facing.
- `nathan`: frontend code authoring or JS/code-editor work; developer-facing.
Create a dedicated employee when the task needs domain-specific behavior, a constrained model set, dedicated custom workflow tools, or a role/persona that should be exposed to business users.
## Avatar Payload Rules
- `avatar` is a preset seed string, not an uploaded file object or external image URL.
- Default seed: `nocobase-015-male`.
- Every new `aiEmployees:create` payload must include `avatar`.
- Before create, validate `avatar` against the supported seeds below. Replace missing, empty, null, or unsupported values with `nocobase-015-male`.
- After create, read back the employee through `aiEmployees:list` or `aiEmployees:listByUser` and verify `avatar` is present.
Supported avatar seeds:
```text
nocobase-001-male
nocobase-002-male
nocobase-003-female
nocobase-004-male
nocobase-005-female
nocobase-006-male
nocobase-007-male
nocobase-008-female
nocobase-009-female
nocobase-010-male
nocobase-011-male
nocobase-012-male
nocobase-013-female
nocobase-014-female
nocobase-015-male
nocobase-016-female
nocobase-017-female
nocobase-018-female
nocobase-019-female
nocobase-020-female
nocobase-021-male
nocobase-022-male
nocobase-023-female
nocobase-024-male
nocobase-025-male
nocobase-026-male
nocobase-027-female
nocobase-028-male
nocobase-029-male
nocobase-030-male
nocobase-031-female
nocobase-032-male
nocobase-033-male
nocobase-034-female
nocobase-035-male
nocobase-036-female
nocobase-037-male
nocobase-038-female
nocobase-039-female
nocobase-040-female
nocobase-041-male
nocobase-042-male
nocobase-043-male
nocobase-044-male
nocobase-045-female
nocobase-046-female
nocobase-047-male
nocobase-048-female
nocobase-049-male
nocobase-050-female
nocobase-051-female
nocobase-052-female
nocobase-053-male
nocobase-054-female
nocobase-055-male
nocobase-056-female
nocobase-057-female
nocobase-058-female
nocobase-059-male
nocobase-060-female
```
## Task Contract Template
Use this internal template before placement:
```json
{
"intent": "what the user wants",
"targetSurface": "page/block/action slot",
"decision": "builtin|js|workflow|ai-employee",
"employee": {
"mode": "reuse|create",
"username": "candidate-or-new-username",
"role": "short business role",
"reason": "why this employee fits"
},
"aiAction": {
"auto": false,
"autoSend": false,
"context": "self|named block|record",
"taskTitle": "short button/task title",
"systemMessage": "stable operational constraints",
"userMessage": "what the employee should do with the block context",
"webSearch": false
}
}
```
## References
- Read `references/ai-employee-api.md` for collections, resource actions, field meanings, and create/update payload notes.
- Read `references/block-action-payload.md` for the public AI employee action shape and placement rules.
- Read `references/examples.md` for reusable classification and payload examples.
Related in Design
contribute
IncludedLocal-only OSS contribution command center. Auto-refreshes the user's in-flight PR and issue state on invoke so conversations start with full context — no need to brief Claude on what's in flight. Helps the user find issues to contribute to on GitHub, builds per-repo dossiers of what each upstream expects (CLA, DCO, branch convention, AI policy, draft-first, review bots, issue templates), runs deterministic gates before any external action so AI-assisted contributions don't reach maintainers as slop. State is markdown-only: candidate files at ~/.contribute-system/candidates/, repo dossiers at ~/.contribute-system/research/, append-only event log at ~/.contribute-system/log.jsonl. No database, no cloud calls. Use when the user asks about their PRs / issues / contributions, wants to find new work to take on, claim an issue, build/refresh a repo's dossier, or draft a Design Issue or PR. Trigger with "/contribute", "what's my PR status", "find a contribution", "claim issue X", "draft a Design Issue for Y", "refresh dossier for Z".
architectural-analysis
IncludedUser-triggered deep architectural analysis of a codebase or scoped subtree across eight modes — information architecture, data flow, integration points, UI surfaces, interaction patterns, data model, control flow, and failure modes. This skill should be used when the user asks to "diagram this codebase," "map the architecture," "show the data flow," "give me an ERD," "trace control flow," "find the integration points," "verify the layout pattern," "audit the UX architecture," or any similar request whose primary deliverable is mermaid diagrams plus cited reports under docs/architecture/. Dispatches haiku/sonnet sub-agents in parallel for per-mode exploration, then verifies every citation mechanically before any node lands in a diagram. Not for one-off prose explanations of code (use code-explanation) or for high-level system design from scratch (use system-design).
mcp
IncludedModel Context Protocol (MCP) server development and tool management. Languages: Python, TypeScript. Capabilities: build MCP servers, integrate external APIs, discover/execute MCP tools, manage multi-server configs, design agent-centric tools. Actions: create, build, integrate, discover, execute, configure MCP servers/tools. Keywords: MCP, Model Context Protocol, MCP server, MCP tool, stdio transport, SSE transport, tool discovery, resource provider, prompt template, external API integration, Gemini CLI MCP, Claude MCP, agent tools, tool execution, server config. Use when: building MCP servers, integrating external APIs as MCP tools, discovering available MCP tools, executing MCP capabilities, configuring multi-server setups, designing tools for AI agents.
react-native-skia
IncludedDesign, build, debug, and optimise high-polish animated graphics in React Native or Expo using @shopify/react-native-skia, Reanimated, and Gesture Handler. Use when the user wants canvas-driven UI, shaders, paths, rich text, image filters, sprite fields, Skottie, video frames, snapshots, web CanvasKit setup, or performance tuning for custom motion-heavy elements such as loaders, hero art, cards, charts, progress indicators, particle systems, or gesture-driven surfaces. Also use when the user asks for fluid, glow, glass, blob, parallax, 60fps/120fps, or GPU-friendly animated effects in React Native, even if they do not explicitly say "Skia". Do not use for ordinary form/layout work with standard views.
plaid
IncludedProduct Led AI Development — guides founders from idea to launched product. Six capabilities: Idea (discover a product idea), Validate (pressure-test the idea against fatal flaws, problem reality, competition, and 2-week MVP feasibility), Plan (vision intake + document generation), Design (translate image references into a design.md spec), Launch (go-to-market strategy), and Build (roadmap execution). Use when someone says "PLAID", "plaid idea", "help me find an idea", "product idea", "idea from my business", "idea from my expertise", "plaid validate", "validate my idea", "pressure-test", "is this idea good", "find fatal flaws", "validate the problem", "plan a product", "define my vision", "generate a PRD", "product strategy", "plaid design", "design from image", "translate image to design", "create design.md", "extract design tokens", "plaid launch", "go-to-market", "launch plan", "GTM strategy", "launch playbook", "plaid build", "build the app", "start building", or "execute the roadmap".
nextjs-framer-motion-animations
IncludedAdds production-safe Motion for React or Framer Motion animations to Next.js apps, including reveal, hover and tap micro-interactions, whileInView, stagger, AnimatePresence, layout and layoutId transitions, reorder, scroll-linked UI, and lightweight route-content transitions. Use when the user asks to add, refactor, or debug Motion or Framer Motion in App Router or Pages Router codebases, especially around server/client boundaries, reduced motion, LazyMotion, bundle size, hydration, or route transitions. Avoid for GSAP-style timelines, WebGL or 3D scenes, heavy scroll storytelling, or CSS-only effects unless Motion is explicitly requested.