MCP Servers with a execute_query tool

Every MCP server Pod knows of that exposes a execute_query tool.

8 servers, ordered by decision readiness and adoption — GitHub stars and npm downloads where available, then reported issue volume. Pod has connected to 7 of them; the rest are publisher-reported and unverified.

Buffer

Buffer is a social media tool that helps you plan, write, and schedule posts across all your channels. The Buffer MCP server connects Buffer to AI assistants like Claude, so you can create posts, save ideas, and check how your content is doing just by asking your AI assistant. No switching tabs, no copy-pasting. It works with all the channels you've already connected to Buffer.

15 tools reported — Live, but requires authorization before it will list tools. Publisher lists 15 tools.

CaliberMind

CaliberMind’s MCP for Claude brings your fully modeled GTM data — pipelines, CRM records, account journeys, and attribution — directly into Claude. All schema context and field relationships flow in automatically, so you can ask any revenue and GTM questions in plain English, run queries, and build dashboards in any environment. Because CaliberMind pre-wires every table relationship, Claude’s answers always match your CRM and GTM numbers exactly for ultimate context-aware revenue intelligence on-demand.

8 tools reported — Live, but requires authorization before it will list tools. Publisher lists 8 tools.

Credible

Ask questions about your company's data in plain language and get real numbers back. Claude retrieves the relevant metrics, definitions, and business rules from the model your team governs in Credible, runs live queries against your data warehouse, and returns results with charts — so "revenue" and "active customer" carry the definitions your team wrote.

What Claude can do with this plugin:

*Retrieve governed context — Claude calls Credible's get_context tool with your question and gets back the relevant sources, dimensions, measures, and business rules from your team's model. Different questions surface different parts of the model; every answer draws from the same governed source.

*Run live warehouse queries — Claude writes queries in Malloy, the open-source data language, and executes them against your warehouse (Snowflake, BigQuery, Databricks, and more). Every number traces back to the modeled definition that produced it.

*Build charts, reports, and dashboards — breakdowns, trends, comparisons, and multi-query reports, rendered in the chat from validated results.

*Apply built-in analysis guidance — bundled skills cover Malloy query patterns, chart selection, and a verification checklist Claude runs before presenting results, plus searchable Malloy and Credible documentation.

*Respect your access controls — view-, row-, and column-level rules are defined in the model and enforced on every query. Claude sees only the workspaces and models you're entitled to.

Answers come from your governed model, so the same question returns the same answer in Claude, your dashboards, and your product. Update a definition in the model and Claude's answers update with it.

Requires a Credible account — get access at credibledata.com

5 tools reported — Live, but requires authorization before it will list tools. Publisher lists 5 tools.

Evarist

Every customer need, hesitation, and buying signal is already in your traffic. Traditional analytics leave your team to find them. Evarist does the analysis for you. Google Analytics, rebuilt for AI.

• Ask anything about your website traffic: What changed this week? Which visitors are closest to buying? Why are fewer visitors becoming customers this month? Which campaigns changed? Where do ready-to-buy visitors leave? Every answer comes with session-level evidence, traceable back to the real journeys behind it.
• Evarist monitors your traffic 24/7 and surfaces what matters before you think to look: traffic sources and campaigns, demand and intent by audience, conversion journeys and funnel drops, revenue linked to each change, and AI traffic from ChatGPT, Claude, Perplexity, and other assistants, measured on whether it creates real business value. And your own agents can act on that intelligence.
• Evarist knows your business (your offer, products, content, and website structure) and reads every session: every page, click, interaction, conversion, and revenue event. Privacy-first: behavioral understanding, not personal identification. One lightweight script, installed once, no analytics configuration project.
• Built on infrastructure processing 60B+ events and 25M+ daily sessions. For marketing teams, e-commerce, SaaS, and lead-generation sites with real conversion questions.

16 tools reported — Live, but requires authorization before it will list tools. Publisher lists 16 tools.

Protobi Claude MCP

Protobi AI Connector allows your AI engine to direct find and query data in your Protobi projects.

4 tools reported — Live, but requires authorization before it will list tools. Publisher lists 4 tools.

Ramp

Enables secure, read-only access to Ramp data for customer-defined logic. Supports loading and querying transactions, reimbursements, bills, purchase orders, cards, vendors, and more, enabling dynamic data fetching and analysis via Claude.

17 tools reported — Live, but requires authorization before it will list tools. Publisher lists 17 tools.

Margins

Most data tools guess. Margins doesn't. Connect 1,000+ sources — Shopify, Amazon, Stripe, Meta, Google, banks, payroll, inventory — and run your business from a Claude conversation on accrual-grade accounting data. Ask anything: sales and profits, blended ROAS, true SKU margin, cash runway, anomalies. Build dashboards and schedule reports in plain English. Every answer is reconciled to the penny — numbers you can defend in a board meeting

9 tools reported — Live, but requires authorization before it will list tools. Publisher lists 9 tools.

Metabase

Search tables and metrics, explore field values, and construct or run queries against your data.

11 tools reported — Publisher-reported; not verified by Pod.

For agents

This page has a Markdown and a JSON twin. Pod is also queryable over MCP at https://api.askpod.ai/mcp/read.