What did Hlido score CentralMind Gateway?

CentralMind Gateway scored 74/100 (STEADY) in Hlido's independent, hands-on review.

Does any vendor pay Hlido for placement?

No. Hlido takes no money from the agents it rates — scoring weights stay private and the evidence behind every verdict is public.

CentralMind Gateway

Name: CentralMind Gateway review
Item: CentralMind Gateway
Rating: 74
Author: Hlido Editor

Infrastructure · tested 2026-08-22 · re-test due 2026-11-22 · by the Hlido desk, not the vendor

In short: An open-source gateway that turns any database into an MCP server or OpenAPI 3.1 in minutes — broad connector coverage and a serious plugin story (PII removal, auth, caching, tracing).

5 PASS · 0 FAIL of 5 public-surface claims

Visit CentralMind Gateway → scorecard.json

Quick answer

CentralMind Gateway scores 74/100 (STEADY) on Hlido’s independent, hands-on test (reviewed 2026-08-22). STEADY (74) for a well-documented, open-source database-to-MCP/OpenAPI gateway with unusually broad connector coverage and a security/observability plugin layer (PII removal, RLS, OAuth, caching, OTel) that takes the act

Hlido Editor · 2026-08-22

CentralMind Gateway addresses a concrete, current problem: safely exposing a production database to AI agents. Point it at a connection string and it generates a REST API (with Swagger UI) and an MCP SSE endpoint agents can consume. The documentation surface is strong and specific — a one-line Docker run, broad connector coverage (Postgres, MySQL, ClickHouse, BigQuery, Snowflake, MongoDB, MS SQL, Oracle, Supabase and more), named integrations (ChatGPT, LangChain, LlamaIndex, Claude Desktop, Cursor), and — importantly for the use case — a plugin layer that speaks directly to the risks of handing an agent a database: PII remover, Presidio anonymizer, API-keys, OAuth, row-level security via Lua, LRU cache and OpenTelemetry. That security/observability plugin set is what lifts it above a thin 'db-to-MCP' wrapper. It's open source (GHCR image, GitHub), which makes it verifiable and self-hostable. What keeps it at STEADY rather than higher from the public surface: it's a young infrastructure project whose reliability, performance and the real robustness of the auto-generated API and RLS enforcement can't be judged from docs alone, and it sits in a fast-moving space where the database-to-MCP pattern is becoming contested.

Why STEADY

STEADY (74) for a well-documented, open-source database-to-MCP/OpenAPI gateway with unusually broad connector coverage and a security/observability plugin layer (PII removal, RLS, OAuth, caching, OTel) that takes the actual risk of the use case seriously. Not VITAL because it's a young infra project whose reliability, performance and the robustness of auto-generated APIs and access controls can't be verified from the documentation surface. Not FADING: current, actively developed, and openly self-hostable.

Public-surface checklist

PASS Homepage loads (required)
PASS Primary value prop (required) — Expose your database to AI agents via MCP or OpenAPI in minutes
PASS Cta present — Quickstart / Docker run
PASS Pricing or access — Open source; GHCR image
PASS Evidence or demo — Docs with runnable quickstart + interactive demo referenced; 1 screenshot captured

What we saw

1 screenshot captured by the Hlido engine during the reviewed run (run-5cd672ff99b57512-centralmind-ai). Our own captures — not vendor marketing material.

CentralMind Gateway — run screenshot 1 (home.png) — home.png

What it does well

Turns a database into both an MCP server and an OpenAPI 3.1 REST API from a connection string
Broad connector coverage: Postgres, MySQL, ClickHouse, BigQuery, Snowflake, MongoDB, MS SQL, Oracle, Supabase and more
Security-aware plugin layer: PII remover, Presidio anonymizer, OAuth, API keys, Lua row-level security
Operable and observable: LRU cache and OpenTelemetry plugins, one-line Docker run, Swagger UI
Open source and self-hostable (GHCR image + GitHub), so behavior is verifiable

What it fails at

Reliability and performance of the auto-generated API can't be judged from the documentation surface
Robustness of access controls (RLS, PII removal) under real adversarial agent use is unproven from docs alone
The database-to-MCP pattern is fast-becoming contested; durable differentiation isn't yet established
Auto-generated APIs risk over-exposing schema/data unless carefully constrained — the burden is on the operator

Red flags

Auto-generating an API/MCP surface over a live database is inherently sensitive; the safety of the result depends heavily on correct plugin configuration (RLS, PII removal, auth). The tooling exists, but the responsibility — and the failure mode — sits with the operator.

Best for

Teams that need to expose an existing database to AI agents via MCP or REST quickly
Data/platform engineers who want PII masking, RLS and OAuth in front of agent access
Self-hosters who want an open-source, Docker-deployable gateway they can audit
Multi-database environments needing one gateway across many connector types

Not recommended for

Teams needing a vendor-backed, SLA'd managed service rather than a self-hosted OSS project
Security-critical exposure where the auto-generated API and access controls must be independently certified first
Simple single-app cases where a purpose-built endpoint is cheaper than a general gateway

Pricing & access

Pricing findable on the public surfacePASS Open source; GHCR image (tested 2026-08-22)

Derived from Hlido-held evidence only (engine checklist + editorial text); quotes are verbatim from the scorecard; not vendor-supplied; re-derived daily. Verify current prices on the vendor's pricing page. Last verified 2026-08-22.

Compared to

Postgrest postgres-rest-vs-multidb-mcp
PostgREST is the mature, Postgres-only way to auto-generate a REST API; CentralMind adds MCP output, many more database connectors, and an agent-oriented security plugin layer. Choose PostgREST for a proven Postgres-only REST layer, CentralMind for multi-DB + MCP + agent controls.

Agent relevance

API CLI MCP Behavioral-testable

Agentic-Commerce Readiness 76/100 · COMMERCE-READY

Independent readiness for agent delegation & transaction. How it’s scored · check live

This product exists to make databases agent-consumable: it emits an MCP SSE endpoint and an OpenAPI 3.1 REST API, with documented integrations for Claude Desktop, Cursor, ChatGPT, LangChain and LlamaIndex. Open source and Docker-deployable, so it is directly behaviorally testable by an evaluating agent.

Agent-friendly score: 9/10

Evidence

Database to MCP server or OpenAPI 3.1 — source (2026-08-22) verified
One-line Docker run generates API + MCP SSE — source (2026-08-22) verified
Broad database connector coverage — source (2026-08-22) verified
Security/observability plugins (PII, OAuth, RLS, cache, OTel) — source (2026-08-22) verified
Open source, GHCR image + GitHub — source (2026-08-22) verified

scorecard.json · transparency passport · registry · methodology

More: compare agents · best of · developer tools · incident registry

Verdict by Hlido Editor, our automated editorial system · Method: public-surface-tier-1+editorial-narrative-v2 · Methodology version 2026.05 · Next review due 2026-11-22

How this page was produced. The scores, claim verdicts and evidence come from automated hands-on testing of the product’s public surface. The written analysis is drafted by an AI system, and pages publish without a person reviewing each one. Hlido publishes this record and answers for it — tell us if anything here is wrong and we will correct it.

Embed this trust badge

Live, always-current independent score — free to embed on your site or README. No vendor pays for placement.

Markdown

[![Hlido trust score](https://hlido.eu/badge/centralmind-gateway.svg)](https://hlido.eu/check/?agent=centralmind-gateway)

HTML

<a href="https://hlido.eu/check/?agent=centralmind-gateway"><img src="https://hlido.eu/badge/centralmind-gateway.svg" alt="Hlido trust score"></a>