Onboarded MCP Server

Onboarded MCP Server

Enables AI assistants to interact with the Onboarded platform through automatic API discovery and execution, with entity memory persistence and optional source code access from local repositories.

Category
访问服务器

README

Onboarded MCP Server

An MCP (Model Context Protocol) server that enables AI assistants like Warp to interact with the Onboarded platform. It provides:

  • API Discovery — Automatically fetch and search OpenAPI specs to find available operations
  • API Execution — Execute any Onboarded API operation with automatic authentication
  • Entity Memory — Persist created entity IDs locally so the agent can reference them later
  • Source Code Access — Optionally read files from your local Onboarded repository

Credentials are stored securely in macOS Keychain and shared with onboarded-cli — tokens never appear in MCP tool calls or responses.

Setup

Prerequisites

Quick Setup (Recommended)

The onboarded CLI handles everything automatically:

onboarded mcp setup

This interactive command will:

  1. Ask where to install (default: ./onboarded-mcp in the current directory)
  2. Clone this repository
  3. Run npm install and npm run build
  4. Add the server to Warp's MCP config (~/.warp/mcp.json)
  5. Optionally configure the Onboarded repo path for repo_read tool

After setup completes:

  1. Restart Warp to load the MCP server
  2. Authenticate: onboarded auth login <profile-name>

Manual Setup

If you prefer to set up manually:

# Clone and build
git clone https://github.com/OnboardedInc/onboarded-mcp.git /ABSOLUTE/PATH/TO/onboarded-mcp
cd /ABSOLUTE/PATH/TO/onboarded-mcp
npm install
npm run build

Add to ~/.warp/mcp.json:

{
  "mcpServers": {
    "onboarded": {
      "command": "node",
      "args": ["/ABSOLUTE/PATH/TO/onboarded-mcp/dist/index.js"]
    }
  }
}

Optionally create ~/.config/onboarded-mcp/config.json for repo_read tool:

{
  "onboardedRepoPath": "/ABSOLUTE/PATH/TO/onboarded"
}

Available Tools

Discovery Tools

Tool Description
spec_sync Fetch and cache OpenAPI specs from Onboarded APIs
ops_search Search for API operations by keyword
ops_describe Get detailed parameter and schema info for an operation

Execution Tools

Tool Description
api_call Execute any API operation by operationId
auth_status Check if credentials exist for a profile

State Tools

Tool Description
state_put Store an entity reference (ID, name, type)
state_query Query stored entities with filters
state_get Get a stored entity by internal or external ID
state_delete Remove an entity from local storage

Repository Tools (optional)

Requires onboardedRepoPath to be configured.

Tool Description
repo_read Read a file from the local Onboarded repository
repo_list List files in a directory

Tool Reference

spec_sync

Fetch and cache an OpenAPI spec.

  • api (required): "v1" or "internal"
  • env: "prod" or "staging" (default: "prod")
  • source: "url" or "cache" (default: "url")

ops_search

Search for API operations by keyword.

  • api (required): "v1" or "internal"
  • query (required): Search term (matches operationId, summary, tags, path)
  • topK: Max results (default: 10)

ops_describe

Get detailed schema for an operation.

  • api (required): "v1" or "internal"
  • operationId (required): The operation to describe

api_call

Execute an API operation.

  • api (required): "v1" or "internal"
  • operationId (required): The operation to call
  • params: Path and query parameters as object
  • body: Request body for POST/PUT/PATCH
  • profile: Auth profile name (default: "default")
  • dryRun: If true, return the request without executing

auth_status

Check authentication status.

  • profile: Profile name (default: "default")

state_put

Store an entity reference.

  • entityType (required): e.g., "employee", "company"
  • api (required): "v1" or "internal"
  • externalId: ID from the Onboarded API
  • name: Human-readable name
  • data: Additional JSON data
  • env: "prod" or "staging" (default: "prod")

state_query

Query stored entities.

  • entityType: Filter by type
  • api: Filter by API
  • env: Filter by environment
  • nameContains: Filter by name substring
  • limit: Max results (default: 100)
  • offset: Pagination offset

state_get

Get a stored entity.

  • id: Internal UUID, or
  • externalId + api + env: Look up by Onboarded API ID

state_delete

Delete a stored entity.

  • id (required): Internal UUID

repo_read

Read a file from the configured repository.

  • path (required): Relative path (e.g., "src/models/employee.py")
  • startLine: Start line number (1-indexed)
  • endLine: End line number (1-indexed)

repo_list

List files in a directory.

  • path: Relative path (default: root)
  • recursive: List recursively (default: false)
  • maxDepth: Max depth for recursion (default: 3)

Data Storage

All data is stored locally and created automatically on first run:

Path Purpose
~/.config/onboarded-mcp/config.json User configuration
~/.cache/onboarded-mcp/ Cached OpenAPI specs
~/.local/share/onboarded-mcp/state.db SQLite database for entity memory

Viewing Stored Entities

# Query all entities
sqlite3 ~/.local/share/onboarded-mcp/state.db "SELECT * FROM entities;"

# Interactive mode
sqlite3 ~/.local/share/onboarded-mcp/state.db

Security

  • Tokens are stored in macOS Keychain, never in files
  • Tokens never appear in MCP tool arguments or responses
  • Authorization headers are redacted in dry-run output
  • Shares credentials with onboarded-cli — authenticate once, use everywhere

Development

npm run build      # Compile TypeScript
npm run dev        # Watch mode
npm run typecheck  # Type checking only

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选