lontar-mcp
MCP server that gives AI agents full blog management over a Lontar headless blogging API, enabling listing, reading, drafting, publishing, updating, and deleting posts.
README
lontar-mcp
MCP server that gives AI agents full blog management over a Lontar headless blogging API — list and read published posts, draft new content, publish, update, and delete.
Lontar is a Laravel package; this server talks to its HTTP API so agents can manage a blog without touching the CMS UI.
Tools
| Tool | Auth | Description |
|---|---|---|
list_posts |
No | List published posts (paginated) |
get_post |
No | Get a published post by slug, including rendered HTML body |
list_drafts |
Yes | List draft posts (paginated) |
create_post |
Yes | Create a new post (draft by default) |
update_post |
Yes | Update an existing post by slug |
delete_post |
Yes | Permanently delete a post |
publish_post |
Yes | Publish a draft (sets published_at to now) |
unpublish_post |
Yes | Unpublish a post (clears published_at) |
list_posts / list_drafts
| Parameter | Type | Default | Description |
|---|---|---|---|
page |
number |
1 |
Page number |
get_post
| Parameter | Type | Description |
|---|---|---|
slug |
string |
Post slug |
create_post
| Parameter | Type | Default | Description |
|---|---|---|---|
title |
string |
— | Post title (slug auto-generated) |
body |
string |
— | Post body (Markdown) |
excerpt |
string |
— | Short excerpt |
published_at |
string |
— | ISO 8601 date; omit to save as draft |
update_post
| Parameter | Type | Description |
|---|---|---|
slug |
string |
Current post slug |
title |
string |
New title (regenerates slug) |
body |
string |
New body (Markdown) |
excerpt |
string | null |
New excerpt |
published_at |
string | null |
New publish date, or null to unpublish |
delete_post / publish_post / unpublish_post
| Parameter | Type | Description |
|---|---|---|
slug |
string |
Post slug |
All tool responses are JSON strings in the MCP text content field.
Install
npm install lontar-mcp
Or clone and build locally:
git clone https://github.com/thesimonharms/lontar-mcp.git
cd lontar-mcp
npm install
npm run build
Requires Node.js 18+.
Environment variables
| Variable | Required | Description |
|---|---|---|
LONTAR_API_URL |
Yes | API base URL, e.g. https://example.com/api |
LONTAR_API_TOKEN |
For write ops | Laravel Sanctum bearer token |
MCP configuration
Cursor / Claude Desktop
{
"mcpServers": {
"lontar": {
"command": "node",
"args": ["C:/absolute/path/to/lontar-mcp/dist/index.js"],
"env": {
"LONTAR_API_URL": "https://your-blog.example.com/api",
"LONTAR_API_TOKEN": "your-sanctum-token"
}
}
}
}
If installed globally or via npx:
{
"mcpServers": {
"lontar": {
"command": "npx",
"args": ["lontar-mcp"],
"env": {
"LONTAR_API_URL": "https://your-blog.example.com/api",
"LONTAR_API_TOKEN": "your-sanctum-token"
}
}
}
}
Examples
List published posts
{ "page": 1 }
Get a post
{ "slug": "hello-world" }
Create a draft
{
"title": "My New Post",
"body": "# Hello\n\nMarkdown content here.",
"excerpt": "A short summary"
}
Publish a draft
{ "slug": "my-new-post" }
Development
npm run build # bundle server to dist/
npm start # run on stdio
npm test # build + run cobasaja tests
Tests live in tests/ and use cobasaja to spawn the server over stdio against an in-memory mock API, asserting tool behaviour end-to-end.
License
MIT © Simon Harms
推荐服务器
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。
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。
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。
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。
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。
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。
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。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
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。