What is MCP?

MCP (Model Context Protocol) is an open standard that lets AI assistants use real tools in other products. Without MCP, an assistant can only tell you how to do something in Kasbly. With MCP, it can look up your posts or save a draft for you, and report back what it did.

Kasbly runs one MCP server, https://mcp.kasbly.com, that any compatible AI app can connect to.

What happens when you connect

The AI app connects to https://mcp.kasbly.com.
You approve access. Kasbly opens in your browser. You sign in, choose a workspace, and pick the permissions the app may use.
The tools appear in the app. The assistant can now call Kasbly tools, such as listing posts or creating a draft, whenever your request needs them.

Why use MCP with Kasbly?

Real actions, not instructions. “Draft three posts for Instagram” creates three drafts in Kasbly, ready for review.
No API keys. You sign in with your Kasbly account. Nothing secret is pasted into the AI app.
Scoped access. Each connection is limited to one workspace and the permissions you approve.
Works with the app you already use. Claude, ChatGPT, Codex, VS Code, Cursor, Gemini CLI and any other client that supports remote MCP servers.

What Kasbly exposes

Area	What the assistant can do
Workspace	Confirm which workspace and permissions it is connected with
Social accounts	List connected accounts and their status
Posts	Find drafts, scheduled, published and failed posts; open one post
Drafts	Save a new draft post for review (never scheduled or published)
Campaigns	List always-on campaigns and open one
Results	Count posts by status for a period; see what works for an account
Workflows	List automation workflows and open one

See the tools reference for every tool and its inputs.

MCP or the API?

Use MCP when a person wants an AI assistant to work in Kasbly through conversation.
Use API keys when you are building your own integration or a backend job that calls Kasbly directly.