Skedyul MCP 伺服器 | PlugMCP
跳到正文 skedyul
The official Node.js SDK for building Skedyul integration apps. Publish tools, webhooks, CRM models, agents, skills, and workflows — then run them on serverless (Lambda) or dedicated (Docker/ECS) compute.
Package: skedyul · Version: 1.4.x · Node: 22+
What you can build
The Skedyul SDK for Node.js
把任意工具接入任意代理。面向非開發者的 MCP 伺服器目錄;閘道承載流量,n8n 承載應用。
產品
目錄
開發者
公司
2026-11-16
Installation npm install skedyul
# or
pnpm add skedyul
Quick start
1. Create skedyul.config.ts import { defineConfig } from 'skedyul'
import pkg from './package.json'
export default defineConfig({
name: 'My Integration',
version: pkg.version,
description: 'What this app does',
computeLayer: 'serverless',
tools: import('./src/registries'),
webhooks: import('./src/registries'),
provision: import('./provision'),
})
2. Aggregate provision config // provision.ts
import type { ProvisionConfig } from 'skedyul'
import env from './env'
import { models, relationships } from './crm'
import * as channels from './channels'
import * as pages from './pages'
import navigation from './pages/navigation'
export default {
env,
navigation,
models: Object.values(models),
channels: Object.values(channels),
pages: Object.values(pages),
relationships,
} satisfies ProvisionConfig
3. Define a tool import { z, type ToolHandler, type ToolDefinition } from 'skedyul'
const inputSchema = z.object({
name: z.string().optional(),
})
type Input = z.infer<typeof inputSchema>
type Output = { message: string }
const handler: ToolHandler<Input, Output> = async (input) => ({
output: { message: `Hello, ${input.name ?? 'world'}!` },
billing: { credits: 1 },
meta: { success: true, message: 'Greeting sent', toolName: 'hello' },
})
export const helloTool: ToolDefinition<Input, Output> = {
name: 'hello',
description: 'Say hello to someone',
inputSchema,
handler,
}
4. Build and run locally skedyul auth login
skedyul dev link --workplace my-clinic
skedyul build
skedyul dev serve --workplace my-clinic
Package exports
Server modes
Serverless (Lambda) import { server } from 'skedyul'
import config from './skedyul.config'
const mcpServer = server.create({
...config,
computeLayer: 'serverless',
tools: toolRegistry,
webhooks: webhookRegistry,
})
export const handler = mcpServer.handler
Or use the dedicated subpath export:
import { handler } from 'skedyul/serverless'
Dedicated (Docker / ECS) const mcpServer = server.create({
...config,
computeLayer: 'dedicated',
defaultPort: 3000,
tools: toolRegistry,
webhooks: webhookRegistry,
})
await mcpServer.listen()
See Server runtime for endpoints, hooks, and compute-layer differences.
Documentation Full documentation lives in docs/ .
Getting started
Platform features
Project structure Recommended layout for a modular integration app:
my-app/
├── skedyul.config.ts # App metadata, registries, build, queues
├── provision.ts # Aggregates version-level config
├── install.ts # Optional per-installation config (shared models)
├── env.ts # Environment variable definitions
├── crm/
│ ├── index.ts
│ ├── relationships.ts
│ └── models/
├── channels/
├── pages/
├── agents/ # Agent YAML v3 files (deployed via CLI)
├── skills/ # Skill YAML files
├── workflows/ # Workflow YAML v2 files
└── src/
├── registries.ts # Tool and webhook registries
└── server.ts # Optional custom server entry
CLI overview # Authentication
skedyul auth login
skedyul auth use <profile>
# Local development (linked mode)
skedyul dev link --workplace <subdomain>
skedyul dev install --workplace <subdomain>
skedyul build
skedyul dev serve --workplace <subdomain>
# Deploy
skedyul dev diff --workplace <subdomain>
skedyul dev deploy --workplace <subdomain>
# Agents, skills, workflows
skedyul agents deploy --file ./agents/booking.yaml --workplace <subdomain>
skedyul skills deploy --file ./skills/scheduling.yaml --workplace <subdomain>
skedyul workflows deploy --file ./workflows/reminder.yaml --workplace <subdomain>
# Testing
skedyul chat --agent booking --workplace <subdomain>
skedyul dev invoke my_tool --workplace <subdomain>
skedyul apps deployment --id dpl_... # build + provision logs
Configuration files
Development (this package) pnpm install
pnpm build # Compile TypeScript + bundle with tsup
pnpm test # Node test runner
Contributing
Use strict TypeScript — no any
Keep MCP transports lean; shared logic belongs in src/server/route-handlers
Add unit tests under tests/ for new behavior
Update docs in docs/ when adding public APIs or CLI commands
Open a PR with a clear summary for review.
License Graphify Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphify skill for Claude Code, Cursor, Codex, and Gemini CLI: local deterministic AST parsing, e Agent Skills Production-grade engineering skills for AI coding agents. Impeccable The design vocabulary for your coding agent. Steer with polish and bolder, review with critique and audit, catch AI slop before it ships. Gemini Cli An open-source AI agent that brings the power of Gemini directly into your terminal. Archify Turn any idea, plan, or codebase into a beautiful interactive diagram. An agent skill for Claude Code, Codex, and more. Filesystem Read, write, search and move files in the directories you allow. The reference filesystem server from the Model Context Protocol project. Everything A test server that exercises every MCP feature: tools, resources, prompts, sampling and notifications. For client builders, from the Model Context Protocol project. Git Read and change local Git repositories: status, diffs, log, commits, branches and checkout. The reference server from the Model Context Protocol project. Chrome Devtools Chrome DevTools for coding agents Open Code Review Secure, fast, efficient, battle-tested at Alibaba's scale. Hybrid architecture code review tool: deterministic pipelines + LLM Agent, precise line-level comments, built-in multi-language ruleset (NPE, Codebase Memory Local code intelligence for developers and AI agents. Connect code, docs and infrastructure in a persistent graph, with CBM Lens, MCP tools, agent integrations and change-aware test selection. Codebase Memory Codebase knowledge graph for AI agents — 162 languages, sub-ms queries, 99% fewer tokens. Developer tools MCP 伺服器 最佳 Developer tools MCP 伺服器 azure-aigateway Configure Azure API Management as an AI Gateway for AI models, MCP tools, and agents. WHEN: semantic caching, token limit, content safety, load balancing, AI model governance, MCP rate limiting, jailb supabase Use when doing ANY task involving Supabase. Triggers: Supabase products (Database, Auth, Edge Functions, Realtime, Storage, Vectors, Cron, Queues); client libraries and SSR integrations (supabase-js,
尚無工具清單。遠端伺服器會在首次健康檢查後取得工具清單。
如何安裝 Skedyul? 在本頁複製適用於 Cursor、Claude Desktop、Claude Code 或 VS Code 的 stdio 設定。它使用發布者的安裝方式,而不是 PlugMCP 代理伺服器。
Skedyul 需要 API 金鑰嗎? 這筆記錄未列出必要的環境變數。
Skedyul 使用哪種傳輸方式? Skedyul 透過 stdio 在本機執行。
Skedyul 上次檢查是什麼時候? 上次檢查時間:2026-11-16。 The Skedyul SDK for Node.js
工具
0 / 10