Back to home
MCP Integration

Connect AI agents to your blog

EZPost supports the Model Context Protocol (MCP), letting AI assistants like Claude, Cursor, Devin, and others read and write your blog posts directly — no HTTP wrappers, no custom code.

What is MCP?

A standard way for AI to talk to your tools

MCP (Model Context Protocol) is an open standard that lets AI assistants discover and call tools automatically. Instead of writing custom API integrations, you just point your AI client at an MCP endpoint — and the AI figures out what tools are available, what inputs they need, and how to use them.

With EZPost's MCP server, your AI agent can list posts, create drafts, publish content, manage categories, and more — all through a single authenticated endpoint.

1

Create an API key

MCP uses the same API key as the External REST API. Create one in your EZPost dashboard:

API Keys page

Go to Settings → API Keys → Add API Key. Choose an expiration period (or Never for permanent keys). Copy the generated key — you won't see it again.

2

Configure your AI client

Each AI client has its own config file. Find yours below, paste the config, and replace YOUR_API_KEY and the URL with your EZPost instance.

Claude Desktop

Config file: ~/Library/Application Support/Claude/claude_desktop_config.json

claude_desktop_config.json
Loading…

Steps:

  1. Open Claude Desktop settings
  2. Click "Developer" → "Edit Config"
  3. Paste the JSON config above (replace URL and API key)
  4. Restart Claude Desktop
  5. Ask Claude: "List my EZPost posts"

Cursor

Config file: .mcp.json (project root)

.mcp.json (project root)
Loading…

Steps:

  1. Create .mcp.json in your project root
  2. Paste the JSON config above
  3. Open Cursor settings → MCP → verify ezpost is connected
  4. Use the chat and ask it to list or create posts

Devin CLI

Config file: .mcp.json (repo root, auto-discovered)

.mcp.json (repo root, auto-discovered)
Loading…

Steps:

  1. Create .mcp.json in your repo root
  2. Set EZPOST_API_KEY environment variable
  3. Devin CLI auto-discovers the config on next session
  4. Ask Devin to manage your posts

Continue.dev / VS Code

Config file: ~/.continue/config.json

config.json
Loading…

Steps:

  1. Open Continue config file
  2. Add the mcpServers section above
  3. Reload VS Code window
  4. Use Continue chat to interact with your posts

Codex

Config file: ~/.codex/config.json

config.json
Loading…

Steps:

  1. Open Codex config file
  2. Add the mcpServers section above
  3. Restart Codex
  4. Ask Codex to read or write your EZPost posts

OpenCode

Config file: opencode.json (project root)

opencode.json (project root)
Loading…

Steps:

  1. Create opencode.json in your project root
  2. Paste the JSON config above
  3. Restart OpenCode
  4. Use the agent to manage your blog posts
3

Available tools

Once connected, your AI agent can call these 10 tools. The agent discovers them automatically — you don't need to memorize anything.

ToolDescriptionKey inputs
list-postsList posts with cursor paginationlimit, cursor, categoryId, search
get-post-by-slugGet a published post by slugslug
create-postCreate a new posttitle, slug, content, description
update-postUpdate a post (partial)id + any field
delete-postDelete a post by IDid
publish-postSet isPublished to trueid
unpublish-postSet isPublished to falseid
list-categoriesList all categories(none)
create-categoryCreate a new categoryname, description
create-subcategoryCreate a subcategorycategoryId, name

Test with curl

Verify your MCP server is working before connecting an AI client:

terminal
Loading…

You should see a JSON response with protocolVersion and serverInfo.name: "ezpost-mcp".

Ready to connect?

Create an API key and start using AI with your blog today.