mcpbeat Sign in

AgentSQL MCP Server

by cuipengdba Your server? Claim it
answering

AgentSQL is answering right now. Last checked moments ago. Last commit 24 Oct 2026.

Guard AI agents' PostgreSQL/MySQL access via MCP: SQL audit, auth, masking, write approval

Uptime history 26 days of history · worst day 88%
26 days agonow
87.0%
Uptime 24h
80 of 92 checks
—
Tools
hidden behind auth
897 ms
Response time
average over 24h
13
Stars
last commit 24 Oct 2026

What the code does

We read the source, 1 h ago · rules 995398eba69f

Evidence

Things with no honest explanation: a promise that contradicts the code, code that runs at install time while hiding what it does, data leaving the machine.

    runuser -u postgres -- "$@"
Capabilities

What this server is able to do. For an MCP server this is often the job itself — a terminal server runs commands because that is what it is for. Listed so you know what you are plugging in, not as an accusation.

    defaultDirectory := filepath.Join("artifacts", "b5-soak", stamp)
var cqlInsert = regexp.MustCompile(`(?i)^INSERT INTO (` + cqlIdent + `) \((` + cqlIdent + `(?:\s*,\s*` + cqlIdent + `)*)\) VALUES \((\?(?:\s*,\s*\?)*)\)$`)
    head = subprocess.check_output(
In the project's build, not in the package

Found in continuous integration, deployment or infrastructure files, or in a neighbouring package of the same monorepo. None of this is installed when you add the server: it describes how the project is built and released. We list it because a leaked key in a build pipeline is still a real problem, but it is not something this server does on your machine.

Is this your server and something here is wrong? Tell us — corrections are free and do not require a plan.

We found things in this code

Code changes quietly between releases, and nobody reads the diff of a dependency. We do, on every release — watch AgentSQL and you get told the day something new turns up.

Three servers free · no card

Connect this server

Endpoint below is the one we actually reach during checks — not the one copied from a README. Last verified 0 min ago.

run in your terminal
claude mcp add agentsql --transport http https://demo.agentsql.cn/mcp
Settings → Connectors → Add custom connector, then paste this address
https://demo.agentsql.cn/mcp
~/.codex/config.toml
[mcp_servers.agentsql]
url = "https://demo.agentsql.cn/mcp"
.cursor/mcp.json
{
  "mcpServers": {
    "agentsql": {
      "url": "https://demo.agentsql.cn/mcp"
    }
  }
}
.vscode/mcp.json
{
  "servers": {
    "agentsql": {
      "type": "http",
      "url": "https://demo.agentsql.cn/mcp"
    }
  }
}

This endpoint answered with an authorization challenge. The server is running, but it did not say what kind of credentials it expects.

Endpoints

URLTransportStateLatencyChecked
https://demo.agentsql.cn/mcp streamable-http needs auth 641 ms 0 min ago

Alternatives to AgentSQL

same job, measured the same way
Jam SQL Studio
by jamsql

Lets AI agents query SQL Server, PostgreSQL, MySQL, Oracle and SQLite via the Jam SQL Studio app.

32 installs/wk local only
Universal DB (read-only)
by alghanim

Read-only, guarded SQL access for AI agents: PostgreSQL, MySQL, ClickHouse, Oracle, SQL Server, Db2

local only
Fastmcp Sqltools
by atarkowska

A MCP server provides SQL database access with support for PostgreSQL, MySQL, and SQLite.

109 installs/wk local only
Mcpolyglot
by ishay60

Policy-checked, audited agent access to Postgres, MySQL, SQLite, MongoDB and OpenAPI REST APIs.

72 installs/wk local only
Rowbase
by djstreet11

Safe, read-only-by-default MySQL, PostgreSQL & SQLite access for AI agents. TOON output.

197 installs/wk local only
DataLink
by pilat

MCP server for secure database access (PostgreSQL, MySQL, SQLite)

31 installs/wk local only
QueryBear
by spencerpauly

Safe, read-only Postgres and MySQL access for AI agents. Audit log + column-level controls.

answering
QueryPilot
by nickklos10

Safe SQL gateway for AI agents: SELECT-only validation, access policies, masking, audit trail, evals

86 installs/wk local only

AgentSQL — questions

Answers built from our own checks of this server.

Why is there no tool list for AgentSQL?
The server answered our handshake with an authorization challenge, so it is running — but it will not describe its tools to an anonymous client. To see them you need to connect with your own credentials. We record it as alive, not as broken: 87.0% of checks in the last 24 hours got a reply.
Is AgentSQL working right now?
We send a real MCP handshake every 15 minutes. Over the last 24 hours 80 of 92 checks got a reply (87.0%), average response time 897 ms. The bar chart above shows every period we have measured.
How do I connect AgentSQL?
Copy the ready config from this page. We generate it for Claude Code, Claude Desktop, Codex, Cursor and VS Code, each in the form that client actually reads. It is a remote server, so there is nothing to install — the client connects to the address. You will need your own credentials: it refuses anonymous clients.
Does AgentSQL need an API key?
Yes. Every time we knock, AgentSQL answers with an authorization challenge instead of its tool list — that is how we know it is running and gated rather than broken. Bring your own credentials and it will talk.
How fast is AgentSQL?
It answers our handshake in 897 ms on average, which is faster than 8% of all working MCP servers we measure. That is on the slow side — worth knowing if the tool sits inside an interactive loop. The comparison comes from our own checks across the whole registry, every 15 minutes.
Is AgentSQL open source?
The code is open source: it is published under the Apache-2.0 licence, written in Go and 13 stars on GitHub. The licence covers the code in the repository, not the hosted service at demo.agentsql.cn, which runs on its operator's own terms.