Kyomi

Connect various data sources to visualize, monitor, and analyze complex datasets through interactive dashboards.

CommunityOnlineNo sign-inGlobalFreeRead-only

What it can do

  • Delete Watch: Permanently delete a watch. This action cannot be undone. Use update_watch to temporarily disable a watch instead.
  • Modify Dashboard: Update an existing dashboard's title and/or content. You must be the owner to update. Use get_chartml_spec to verify ChartML syntax before updating. ChartML blocks are validated befo
  • Save Learning: Save knowledge about how to navigate the data warehouse. NOT for saving analysis results or business insights. IMPORTANT: If your learning mentions specific tables or columns, you MUST:

What data it sees

Do you need an account

No: the server works without sign-in

Connect various data sources to visualize, monitor, and analyze complex datasets through interactive dashboards. Automate data monitoring with scheduled reports and alerts that notify you of significant changes or trends. Leverage advanced forecasting and semantic search to uncover hidden insights and predict future business outcomes.

How to connect

How to connect Kyomi to Claude

The server address to paste into Claude settings:

https://kyomi-mcp--kyomi.run.tools

  1. Open Claude (claude.ai or the desktop app).
  2. Go to Settings → Connectors.
  3. Click Add custom connector.
  4. Paste the server address copied below into Remote MCP server URL and click Add.
  5. In a chat, click + → Connectors and switch the new connector on.

Custom connectors are available on Free, Pro, Max, Team and Enterprise plans (Free is limited to one). On Team and Enterprise an organization Owner adds the connector first under Organization settings → Connectors.

Authorization

After you click Add, a sign-in window for the service opens. Sign in with your own account and approve access. Claude never sees your password.

How to connect Kyomi to ChatGPT

The server address to paste into ChatGPT settings:

https://kyomi-mcp--kyomi.run.tools

  1. Open ChatGPT in a browser (chatgpt.com). A Plus, Pro, Business, Enterprise or Edu plan is required.
  2. Turn on developer mode once: Settings → Apps → Advanced settings → Developer mode.
  3. Open Settings → Connectors and click Create.
  4. Fill in the form: Name (anything), Description (one line about what the service does), MCP server URL (copy it below).
  5. Under Authentication choose OAuth if the service requires sign-in, otherwise None. Click Create.
  6. In a new chat open + → Apps/Connectors and enable the connector.

OpenAI has renamed this section before (Connectors → Apps/Plugins). If the label differs, search settings for "developer mode". On Business/Enterprise workspaces an admin must allow custom connectors first.

Authorization

On first use ChatGPT opens the service's sign-in window. Sign in and approve access.

How to connect Kyomi to Cursor

The server address to paste into Cursor settings:

https://kyomi-mcp--kyomi.run.tools

Fastest: click Open in Cursor below and confirm the prompt.

Manually:

  1. In Cursor open Settings → Cursor Settings → MCP and click Add new global MCP server.
  2. Paste the JSON copied below into ~/.cursor/mcp.json (per project: .cursor/mcp.json in the repo root).
  3. Save the file. The server appears in the MCP list; authorize it there if asked.

Authorization

If the service needs sign-in, an authorize button appears next to the server in the MCP list.

Server tool list (27)

Raw names from tools/list. Only developers need these.

delete_watchPermanently delete a watch. This action cannot be undone. Use update_watch to temporarily disable a watch instead.
modify_dashboardUpdate an existing dashboard's title and/or content. You must be the owner to update. Use get_chartml_spec to verify ChartML syntax before updating. ChartML blocks are validated before saving - invalid syntax will return detailed errors.
save_learningSave knowledge about how to navigate the data warehouse. NOT for saving analysis results or business insights. IMPORTANT: If your learning mentions specific tables or columns, you MUST: (1) verify they exist using get_table_info first, (2) include related_tables with fully qualified names (e.g. ["public.orders"]), (3) include related_columns for key columns (e.g. ["public.orders.total_amount"]). Learnings referencing tables without this metadata will be rejected.
list_datasourcesList all available datasources in the current workspace. Returns each datasource's slug, name, type, and number of indexed tables.
forecast_dataRun time series forecasting on data from a datasource. Returns forecast predictions with confidence intervals that you can use to make projections and recommendations. Use this when the user asks about future trends, projections, or predictions. The tool queries the datasource, fits a forecasting model, and returns predicted values with upper/lower confidence bounds. Models: 'auto' (recommended - selects best model via cross-validation), 'ets' (exponential smoothing), 'linear', 'exponential' (growth curves), 'logistic' (S-curves/saturation). Minimum 4 data points required. Always include ORDER BY on the timestamp column in your SQL query.
browse_resourcesList available Kyomi documentation resources. Returns a catalog of docs:// URIs with titles and descriptions. Use this to discover what documentation is available, then use read_resource to read specific docs.
get_workspace_infoGet information about the current workspace including all members and their emails.
get_watch_infoGet detailed information for a specific watch including recent execution history. Use this to view configuration, debug watch behavior, check why a watch didn't alert, or verify it is running on schedule.
search_dashboardsSearch for dashboards in the current workspace. Supports sorting by popularity, recency, or creation date. Use this to find existing dashboards, check for duplicates before creating, or help users locate specific dashboards. Can return top 10 most popular dashboards.
render_chartRender a ChartML visualization as an interactive chart or PNG image. Accepts a ChartML specification as YAML or JSON string. Resolves data queries and produces the final visualization. Use get_chartml_spec first to understand ChartML syntax.
get_dashboard_infoGet detailed information about a specific dashboard including full content.
update_analytics_siteUpdate an analytics site's name, allowed domains, and/or datasource slug. If domains change, the tracking snippet is regenerated (the old snippet stops working).
read_resourceRead a specific Kyomi documentation resource by its docs:// URI. Returns the full markdown content of the document. Use browse_resources first to discover available URIs.
update_watchUpdate an existing Kyomi Watch. You can change name, prompt, schedule, mode, enabled status, Slack channel, or email alert settings. Only include parameters you want to change.
list_analytics_sitesList all analytics sites in the current workspace. Returns site details including tracking snippets and linked datasource slugs.
create_dashboardCreate a new dashboard. You MUST search for existing dashboards first using search_dashboards to avoid duplicates. Use get_chartml_spec to check ChartML syntax before creating dashboards with charts. ChartML blocks are validated before saving - invalid syntax will return detailed errors.
create_watchCreate a Kyomi Watch - automated data monitoring that runs on a schedule. **Two Modes (REQUIRED - you must choose one):** 1. **Alert Mode** (mode="alert") - Conditional monitoring 2. **Report Mode** (mode="report") - Scheduled reports **Workflow:** 1. Use search_watches to check for duplicates first 2. Explore data schema 3. Set verified_no_duplicates=true after checking **Schedule:** Cron expression in UTC (5 fields: min hour day month weekday) NOTE: Premium feature (Pro/Team plans).
search_knowledgeSearch the workspace's knowledge base for relevant tables, learnings, and metrics using semantic search. Use this to discover what data is available before querying.
search_watchesSearch for existing watches to avoid duplicates or find watches to modify. Leave the query empty to list all watches. Returns matching watches with full details (name, prompt, schedule, mode, queries, status, last execution).
get_chartml_specGet ChartML specification documentation for creating visualizations.
validate_sqlValidate SQL query syntax without executing. Returns validation status, error message with line/column if invalid.
query_datasourceExecute a SQL query against any datasource and return results. HARD LIMIT: 20 rows maximum. Use fully qualified table names.
get_table_infoGet detailed information about a table from any datasource. Returns schema, column names, types, descriptions, and metadata.
trigger_watchManually trigger a watch to run immediately. The watch must be enabled. Use get_watch_info afterwards to see the results.
create_analytics_siteCreate an analytics site for tracking website events. Returns a <script> snippet to embed in the site's HTML <head>. Automatically provisions a queryable ClickHouse datasource that you can use with query_datasource, create_dashboard, and other tools.
delete_analytics_siteDelete an analytics site and its auto-provisioned datasource. This removes the ClickHouse user, row policy, and all associated dashboards/watches. This action cannot be undone.
delete_dashboardDelete a dashboard. You must be the owner to delete. This action cannot be undone.