by thoughtspot
Provides secure OAuth authentication and a suite of tools for querying and retrieving data from ThoughtSpot instances via a remote server hosted on Cloudflare.
ThoughtSpot MCP Server enables remote, OAuth‑based access to ThoughtSpot analytics. It exposes MCP tools such as ping, getRelevantQuestions, getAnswer, createLiveboard, and datasource suggestions, allowing LLMs and client applications to retrieve and act on business data.
https://agent.thoughtspot.app/mcp (or the OpenAI‑specific endpoint).{
"mcpServers": {
"ThoughtSpot": {
"command": "npx",
"args": [
"mcp-remote",
"https://agent.thoughtspot.app/mcp"
]
}
}
}
TS_AUTH_TOKEN via ThoughtSpot’s REST Playground or the /api/rest/2.0/auth/session/token endpoint and provide it in the request headers.ping, getRelevantQuestions, getAnswer, createLiveboard, getDataSourceSuggestions.@thoughtspot/mcp-server npm package for restricted networks.bearer endpoints to call MCP tools from OpenAI or Anthropic APIs without full OAuth flows.Q: How do I get a TS_AUTH_TOKEN?
A: Generate it in ThoughtSpot’s Develop → Rest Playground (Full access token) or call https://<your‑instance>/api/rest/2.0/auth/session/token with your credentials.
Q: What if I encounter CORS or SAML errors?
A: Add agent.thoughtspot.app to the CORS whitelist and SAML redirect domains in ThoughtSpot’s security settings.
Q: Can I run the server on my own infrastructure?
A: Yes, pull the Docker image from the repository and follow the instructions in deploy/README.md.
Q: Which Node version is required? A: Node ≥ 18.
Q: My authentication keeps failing with stale tokens.
A: Remove the local cache ~/.mcp-auth and re‑authenticate.
The ThoughtSpot MCP Server provides secure OAuth-based authentication and a set of tools for querying and retrieving relevant data from your ThoughtSpot instance. It's a remote server hosted on Cloudflare.
If you do not have a Thoughtspot account, create one for free here.
Learn more about ThoughtSpot.
Join our Discord to get support.
If using a client which supports remote MCPs natively (Claude.ai etc) then just enter:
MCP Server URL:
https://agent.thoughtspot.app/mcp
https://agent.thoughtspot.app/openai/mcp
To configure this MCP server in your MCP client (such as Claude Desktop, Windsurf, Cursor, etc.) which do not support remote MCPs, add the following configuration to your MCP client settings:
{
"mcpServers": {
"ThoughtSpot": {
"command": "npx",
"args": [
"mcp-remote",
"https://agent.thoughtspot.app/mcp"
]
}
}
}
See the Troubleshooting section for any errors / more details.
See the video below for a complete demo.
Here is a demo video using Claude Desktop.
https://github.com/user-attachments/assets/72a5383a-7b2a-4987-857a-b6218d7eea22
Watch on Loom
ThoughtSpot's remote MCP server can be used in LLM APIs which support calling MCP tools.
Here are examples with the common LLM providers:
curl https://api.openai.com/v1/responses \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $OPENAI_API_KEY" \
-d '{
"model": "gpt-4.1",
"tools": [
{
"type": "mcp",
"server_label": "thoughtspot",
"server_url": "https://agent.thoughtspot.app/bearer/mcp",
"headers": {
"Authorization": "Bearer $TS_AUTH_TOKEN",
"x-ts-host": "my-thoughtspot-instance.thoughtspot.cloud"
}
}
],
"input": "How can I increase my sales ?"
}'
More details on how can you use OpenAI API with MCP tool calling can be found here.
curl https://api.anthropic.com/v1/messages \
-H "Content-Type: application/json" \
-H "X-API-Key: $ANTHROPIC_API_KEY" \
-H "anthropic-version: 2023-06-01" \
-H "anthropic-beta: mcp-client-2025-04-04" \
-d '{
"model": "claude-sonnet-4-20250514",
"max_tokens": 1000,
"messages": [{
"role": "user",
"content": "How do I increase my sales ?"
}],
"mcp_servers": [
{
"type": "url",
"url": "https://agent.thoughtspot.app/bearer/mcp",
"name": "thoughtspot",
"authorization_token": "$TS_AUTH_TOKEN@my-thoughtspot-instance.thoughtspot.cloud"
}
]
}'
Note: In the authorization_token field we have suffixed the ThoughtSpot instance host as well with the @ symbol to the TS_AUTH_TOKEN.
More details on Claude MCP connector here.
For API usage, you would the token endpoints with a secret_key to generate the API_TOKEN for a specific user/role, more details here.
ping: Test connectivity and authentication.getRelevantQuestions: Get relevant data questions from ThoughtSpot analytics based on a user query.getAnswer: Get the answer to a specific question from ThoughtSpot analytics.createLiveboard: Create a liveboard from a list of answers.getDataSourceSuggestions: Get datasource suggestions for a given query.datasources: List of ThoughtSpot Data models the user has access to.Use the published docker image to deploy the MCP server in your own environment.
See this for details.
If you are unable to use the remote MCP server due to connectivity restrictions on your Thoughtspot instance. You could use the stdio local transport using the npm package.
Here is how to configure stdio with MCP Client:
{
"mcpServers": {
"ThoughtSpot": {
"command": "npx",
"args": [
"@thoughtspot/mcp-server"
],
"env": {
"TS_INSTANCE": "<your Thoughtspot Instance URL>",
"TS_AUTH_TOKEN": "<ThoughtSpot Access Token>"
}
}
}
}
TS_AUTH_TOKEN ?TS_AUTH_TOKENOauth errors due to CORS/SAML.
Make sure to add the following entries in your ThoughtSpot instance:
CORS
SAML (need to be Admin)
MCP server install error due to node issues
node -v500 error from MCP server
Stale MCP auth
rm -rf ~/.mcp-auth.npm install
.dev.vars and fill in your ThoughtSpot instance URL and access token.npm run dev
/mcp: MCP HTTP Streaming endpoint/sse: Server-sent events for MCP/api: MCP tools exposed as HTTP endpoints/authorize, /token, /register: OAuth endpoints/bearer/mcp, /bearer/sse: MCP endpoints as bearer auth instead of Oauth, mainly for use in APIs or in cases where Oauth is not working.MCP Server, © ThoughtSpot, Inc. 2025
Please log in to share your review and rating for this MCP.
{
"mcpServers": {
"ThoughtSpot": {
"command": "npx",
"args": [
"mcp-remote",
"https://agent.thoughtspot.app/mcp"
],
"env": {}
}
}
}claude mcp add ThoughtSpot npx mcp-remote https://agent.thoughtspot.app/mcpExplore related MCPs that share similar capabilities and solve comparable challenges
by mindsdb
Enables humans, AI agents, and applications to retrieve highly accurate answers across large‑scale data sources, unifying heterogeneous databases, warehouses, and SaaS platforms.
by mckinsey
Build high-quality data visualization apps quickly using a low-code toolkit that leverages Plotly, Dash, and Pydantic.
by antvis
Offers over 25 AntV chart types for automated chart generation and data analysis, callable via MCP tools, CLI, HTTP, SSE, or streamable transports.
by reading-plus-ai
A versatile tool that enables interactive data exploration through prompts, CSV loading, and script execution.
by Canner
Provides a semantic engine that lets MCP clients and AI agents query enterprise data with contextual understanding, precise calculations, and built‑in governance.
by surendranb
Provides natural‑language access to Google Analytics 4 data via MCP, exposing over 200 dimensions and metrics for Claude, Cursor and other compatible clients.
by ergut
Provides secure, read‑only access to BigQuery datasets, allowing large language models to query and analyze data through a standardized interface.
by isaacwasserman
Provides an interface for LLMs to visualize data using Vega‑Lite syntax, supporting saving of data tables and rendering visualizations as either a full Vega‑Lite specification (text) or a base64‑encoded PNG image.
by vantage-sh
Fetch and explore cloud cost and usage data from a Vantage account using natural language through AI assistants and MCP clients.