Elasticflow

What Elasticflow MCP server does

The Elasticflow MCP server gives AI clients access to shared workspaces used by teams, projects, or departments. Agents can inspect workspace membership, settings, and usage, then create or update workspaces when their account permissions allow it.

Structured data operations cover the full table lifecycle. An agent can list tables, inspect schemas and row counts, review available column types, create or remove tables and columns, query rows with filters, update cell values, and add or delete records. Row-linking tools connect related records across tables, such as contacts and deals.

Document tools support both discovery and editing. Agents can list, read, search, create, and update documents, make targeted section changes, append new material, and manage embedded images and files. The README also describes file tools for uploading and reading files, plus interface tools for creating and publishing dashboards, client portals, and internal applications.

How it works

The Elasticflow MCP server uses Streamable HTTP. The main endpoint exposes the complete tool set at https://mcp.elasticflow.app/mcp. Smaller endpoints are available when a client or workflow needs only part of the platform: /data covers workspaces and tables, /content covers documents and files, and /platform covers dashboards and publishing. These endpoints use the same account and authenticate independently.

Authentication uses OAuth 2.1 with PKCE. On first connection, the client opens a browser so the user can sign in or create an Elasticflow account. Access is controlled through scopes for workspaces, tables, documents, files, and interfaces.

Setup and configuration

Add the Streamable HTTP URL to an MCP-compatible client. Claude Desktop, Cursor, and VS Code use JSON configuration with the server URL; Claude Code can add it with the streamable-http transport. The full endpoint is:

https://mcp.elasticflow.app/mcp

No environment variables, API keys, or manually copied tokens are required according to the provided setup instructions. An Elasticflow account is required.

Tools and capabilities

• Manage workspaces, including creation, settings, membership details, and usage views.
• Create, inspect, query, modify, and delete tables, columns, rows, and cross-table links.
• Search, write, patch, append, and organize documents and their embedded assets.
• Upload, inspect, and read files through the content-oriented tools.
• Create, configure, build, publish, and roll back dashboards or client portals through interface tools.

Limitations and notes

The full endpoint exposes 51 tools, while the supplied tool list details workspace, table, and document operations. The README separately documents file and interface tool groups. Clients with tool-count limits can use the narrower data, content, or platform endpoints instead of installing every capability. Available actions still depend on the OAuth scopes and permissions granted to the connected Elasticflow account.