Blogger MCP
A TypeScript MCP server for managing Blogger drafts and posts, project devlogs with screenshots, and Strava running retrospectives via natural language.
README
Blogger MCP
TypeScript MCP server for Blogger drafts, post management, project devlogs, and Strava running retrospectives.
한국어 안내
Blogger MCP는 Codex, Claude 같은 MCP 클라이언트에서 Blogger 글을 작성하고 관리하기 위한 TypeScript 기반 MCP 서버입니다. 기본 동작은 항상 초안 생성이며, 실제 게시는 publish_blogger_post를 명시적으로 호출할 때만 실행됩니다.
주요 기능:
- Blogger 글 초안 생성
- 최근 글/초안 목록 조회
- 기존 글 수정
- 명시적 게시
- Google OAuth refresh token 로컬 저장
- Markdown을 HTML로 변환해 Blogger 본문에 삽입
- 프로젝트 개발일지 초안 생성
- Strava 활동 데이터 기반 러닝 회고 초안 생성
빠른 시작
git clone https://github.com/powerofdeen/BloggerMCP.git
cd BloggerMCP
npm install
cp .env.example .env
npm run build
.env에 Blogger용 Google OAuth 값과 블로그 ID를 입력합니다.
GOOGLE_CLIENT_ID=...
GOOGLE_CLIENT_SECRET=...
GOOGLE_REDIRECT_URI=http://localhost:3000/oauth2callback
BLOGGER_BLOG_ID=...
Google OAuth 로그인을 진행합니다.
npm run oauth:google
터미널에 출력된 Google 로그인 URL을 열고 승인하면, http://localhost:3000/oauth2callback에서 refresh token이 자동으로 저장됩니다. 토큰은 .tokens/google.json에 저장되며 Git에는 포함되지 않습니다.
Codex에 등록
아래 /absolute/path/to/BloggerMCP는 git clone으로 내려받은 BloggerMCP 폴더의 절대경로로 바꿔야 합니다. 프로젝트 폴더 안에서 pwd를 실행하면 현재 절대경로를 확인할 수 있습니다.
pwd
codex mcp add blogger -- /opt/homebrew/bin/npm --prefix /absolute/path/to/BloggerMCP run start
등록 확인:
codex mcp list
Codex에서 /mcp를 입력해도 blogger 서버가 보이는지 확인할 수 있습니다.
Claude Desktop에 등록
macOS 기준 Claude Desktop 설정 파일:
~/Library/Application Support/Claude/claude_desktop_config.json
설정 예시:
아래 /absolute/path/to/BloggerMCP는 git clone으로 내려받은 BloggerMCP 폴더의 절대경로입니다. 예를 들어 pwd 결과가 /Users/me/dev/BloggerMCP라면 그 값을 넣으면 됩니다.
{
"mcpServers": {
"blogger": {
"command": "/opt/homebrew/bin/npm",
"args": [
"--prefix",
"/absolute/path/to/BloggerMCP",
"run",
"start"
]
}
}
}
저장 후 Claude Desktop을 완전히 종료했다가 다시 실행하면 됩니다.
사용 예시
초안 생성:
create_blogger_draft로 제목은 "테스트 글", 본문은 "## 테스트\nBlogger MCP로 만든 첫 초안입니다."로 초안 만들어줘
최근 글/초안 조회:
list_blogger_posts로 최근 글과 초안 보여줘
프로젝트 개발일지 초안 생성:
create_project_devlog_draft로 프로젝트명 BloggerMCP, 오늘 작업 내용은 "Blogger MCP 서버 구현", "OAuth callback helper 추가"로 초안 만들어줘
게시:
publish_blogger_post로 postId="POST_ID" confirmPublish=true 실행해줘
제공 도구
blogger_login: Google OAuth 로그인 URL 생성 또는 OAuth code 교환create_blogger_draft: Markdown 기반 Blogger 초안 생성list_blogger_posts: 최근 글/초안 목록 조회update_blogger_post: 기존 Blogger 글 수정publish_blogger_post: 명시적 게시score_blog_post: 글 구조와 SEO 기본 점수 확인create_project_devlog_draft: 프로젝트 개발일지 초안 생성create_strava_activity_draft: Strava 활동 기반 러닝 회고 초안 생성
안전 원칙
- 기본은 항상 초안 생성입니다.
- 게시하려면
publish_blogger_post를 명시적으로 호출해야 합니다. publish_blogger_post에는confirmPublish: true가 필요합니다.- 이미지는 Blogger 편집기에서 직접 업로드하는 흐름을 권장합니다.
.env,.tokens는 Git에 올리지 않습니다.- client secret, access token, refresh token은 로그나 응답에 출력하지 않습니다.
Features
- Create Blogger posts as drafts by default.
- List recent live posts and drafts.
- Update existing posts without publishing.
- Publish only through
publish_blogger_postwithconfirmPublish: true. - Store Google OAuth refresh tokens locally.
- Convert Markdown input to Blogger HTML.
- Separate Strava OAuth/client module for running retrospective drafts.
- No client secret, access token, or refresh token is logged or returned.
Setup
npm install
cp .env.example .env
npm run build
Fill .env:
GOOGLE_CLIENT_ID=...
GOOGLE_CLIENT_SECRET=...
GOOGLE_REDIRECT_URI=http://localhost:3000/oauth2callback
BLOGGER_BLOG_ID=...
STRAVA_CLIENT_ID=...
STRAVA_CLIENT_SECRET=...
STRAVA_REDIRECT_URI=http://localhost:3000/strava/callback
Google Cloud / Blogger API
- Open Google Cloud Console.
- Create or select a project.
- Enable Blogger API v3.
- Configure OAuth consent screen.
- Create an OAuth client ID.
- Add the redirect URI from
.env, for examplehttp://localhost:3000/oauth2callback. - Copy client ID and client secret into
.env. - Find your Blogger blog ID in Blogger settings or via Blogger API and set
BLOGGER_BLOG_ID.
First login:
Recommended local callback flow:
npm run oauth:google
Then open the printed Google OAuth URL. After approval, the local
http://localhost:3000/oauth2callback page stores the refresh token and shows a
success message.
Manual fallback:
- Call MCP tool
blogger_loginwith nocode. - Open the returned
authUrl. - Copy the
codequery parameter from the redirect URL. - Call
blogger_loginagain with thatcode.
The refresh token is stored under .tokens/google.json.
Strava Setup
- Create an app at Strava API Settings.
- Set the callback domain/redirect to match
STRAVA_REDIRECT_URI. - Put
STRAVA_CLIENT_IDandSTRAVA_CLIENT_SECRETin.env. - Call
create_strava_activity_draftwithmode: "auth_url". - Open the URL, approve access, then call the tool with
mode: "exchange_code"and the returned code.
The Strava token is stored under .tokens/strava.json.
MCP Configuration
In the examples below, replace /absolute/path/to/BloggerMCP with the absolute
path of the local folder created by git clone. Run pwd inside the
BloggerMCP directory to find it.
Example:
{
"mcpServers": {
"blogger": {
"command": "node",
"args": ["/absolute/path/to/BloggerMCP/dist/server.js"],
"env": {
"NODE_ENV": "production"
}
}
}
}
For local development:
{
"mcpServers": {
"blogger-dev": {
"command": "npm",
"args": ["run", "dev"],
"cwd": "/absolute/path/to/BloggerMCP"
}
}
}
Tools
blogger_login
Returns a Google OAuth URL, or exchanges an OAuth code and stores a refresh token.
create_blogger_draft
Creates a draft post from Markdown.
Inputs include:
titlemarkdownlabels
Local image upload is intentionally not part of the main workflow. Create the draft with MCP, then upload and place images in the Blogger editor.
list_blogger_posts
Lists recent posts. Use status: ["draft"] for drafts only.
update_blogger_post
Patches title, content, and labels. It does not publish.
publish_blogger_post
Publishes a post only when called explicitly:
{
"postId": "123",
"confirmPublish": true
}
score_blog_post
Scores title length, heading structure, content depth, wrap-up, and SEO labels.
create_project_devlog_draft
Creates a structured devlog draft:
- 오늘 한 일
- 문제 상황
- 해결 방법
- 결과 화면
- 배운 점
- 다음 작업
create_strava_activity_draft
Modes:
auth_urlexchange_codelist_recentcreate_draft
The running retrospective includes distance, time, average pace, average heart rate, elevation, activity name, date, and lap pace data when available.
Image Workflow
Blogger's web editor can upload images directly into Blogger/Google-hosted storage, but the Blogger API draft flow used by this MCP expects post content HTML and does not provide the same simple local-file upload path.
Recommended workflow:
- Use MCP to create or update the text draft.
- Open the draft in Blogger.
- Upload screenshots or attached images with the Blogger editor.
- Review the final layout and publish manually or with
publish_blogger_post.
Safety Notes
- Draft creation is the default.
- Nothing publishes unless
publish_blogger_postis explicitly called. publish_blogger_postrequiresconfirmPublish: true.- Client secrets and tokens must stay in
.envand.tokens. .envand.tokensare ignored by git.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.