What can Supabase MCP Server do?

mcp-standardized supabase api tool registration and schema exposure, postgresql database query execution with postgrest api abstraction, shared utility abstractions for mcp server development, oauth 2.1 dynamic client registration for secure credential exchange, monorepo workspace management with pnpm and biome tooling, mock system for testing mcp servers without live api access, supabase management api operations for project administration, database branching and preview deployment orchestration, authentication user and session management through auth api, cloud storage bucket and object management with signed urls, edge function invocation and deployment management, multi-transport mcp server with http, stdio, and sse support, cost tracking and usage analytics through management api, feature group-based capability control and safety gating

Supabase MCP Server

MCP ServerFree

Manage Supabase databases, auth, and storage via MCP.

Open Source

/ 100

14 capabilities

Capabilities14 decomposed

mcp-standardized supabase api tool registration and schema exposure

Medium confidence

Implements the Model Context Protocol specification to expose Supabase operations as discoverable tools with JSON Schema definitions. The server uses a platform abstraction layer that maps Supabase Management API endpoints to MCP tool schemas, enabling any MCP-compatible client (Claude Desktop, Cursor, VS Code Copilot, Windsurf) to dynamically discover and invoke Supabase operations without custom integration code. Tool definitions are generated from the platform interface and include input/output schemas for type-safe invocation.

Solves for

Connect an AI assistant to my Supabase project without writing custom API integrationsEnable Claude or Cursor to discover what Supabase operations are available in my projectLet an LLM invoke database queries, manage auth users, and trigger edge functions through standardized tool calling

Best for

AI assistant users (Claude, Cursor, Windsurf) wanting native Supabase integration

Teams building LLM agents that need backend-as-a-service capabilities

Developers migrating from custom REST integrations to standardized MCP protocol

Requires

MCP-compatible client (Claude Desktop, Cursor, VS Code Copilot, or Windsurf)

Supabase project with valid API credentials (project URL and API key)

Network access to Supabase Management API or local Supabase instance

Limitations

Requires MCP-compatible client — not usable with standard REST API clients

Tool discovery is static per server initialization — dynamic tool registration not supported

Schema generation depends on Supabase API stability — breaking API changes require server updates

What makes it unique

Uses a platform abstraction layer (Platform interface + API Platform implementation) that decouples MCP tool definitions from underlying Supabase API changes, enabling single-source-of-truth tool registration across multiple transport mechanisms (HTTP, stdio, SSE). This differs from monolithic integrations that hardcode API mappings.

vs alternatives

Provides protocol-standard tool exposure vs proprietary integrations, enabling use across any MCP-compatible client without vendor lock-in or duplicate integration work

postgresql database query execution with postgrest api abstraction

Medium confidence

Executes SQL queries and performs CRUD operations against Supabase PostgreSQL databases through a dedicated PostgREST MCP server (@supabase/mcp-server-postgrest). The server translates MCP tool invocations into PostgREST HTTP requests, handling query parameter construction, response parsing, and error translation. Supports parameterized queries to prevent SQL injection and returns results as structured JSON with metadata about affected rows and query execution.

Solves for

Run SELECT, INSERT, UPDATE, DELETE queries on my Supabase database from an AI assistantLet Claude or Cursor perform data analysis by querying my database schemaExecute parameterized queries safely without exposing raw SQL to the LLM

Best for

Developers building LLM agents that need read/write database access

Data analysts using AI assistants to explore and transform Supabase data

Teams automating database operations through AI-driven workflows

Requires

Supabase project with PostgREST API enabled

Valid database credentials with appropriate table-level permissions

@supabase/mcp-server-postgrest package installed and configured

Limitations

PostgREST API limitations apply — complex joins and aggregations may require raw SQL

Row-level security (RLS) policies are enforced — LLM cannot bypass database-level access controls

Large result sets may exceed MCP message size limits — pagination required for bulk queries

What makes it unique

Separates PostgREST operations into a dedicated MCP server package, allowing independent scaling and feature development. Uses parameter binding to prevent SQL injection while maintaining query expressiveness, and integrates RLS enforcement at the database layer rather than application layer.

vs alternatives

Safer than raw SQL execution because it enforces Supabase RLS policies and uses parameterized queries, while more flexible than ORM-based approaches because it supports arbitrary SQL and PostgREST filtering syntax

shared utility abstractions for mcp server development

Medium confidence

Provides reusable TypeScript utilities and abstractions (@supabase/mcp-utils package) for building MCP servers, including platform interface definitions, tool schema builders, and common patterns for API integration. The utilities package defines the Platform interface that abstracts Supabase API differences, allowing multiple MCP server implementations (Management API, PostgREST) to share common patterns. Includes mock system for testing and development without live API access.

Solves for

Build custom MCP servers for Supabase or other services using shared patternsExtend Supabase MCP with additional tools without duplicating core infrastructureTest MCP server implementations without requiring live API credentialsStandardize tool schema definitions across multiple MCP server packages

Best for

MCP server developers building Supabase integrations

Teams extending Supabase MCP with custom tools

Contributors to the Supabase MCP project

Requires

TypeScript 5.0+ and Node.js 18+

@supabase/mcp-utils package installed

Understanding of MCP protocol and Supabase API patterns

Limitations

Utilities are TypeScript-only — no Python or other language support

Mock system is limited to predefined scenarios — custom mocking requires code changes

Abstractions add complexity for simple use cases — overhead not justified for single-tool servers

What makes it unique

Provides platform abstraction pattern that allows multiple MCP server implementations to share common infrastructure while supporting different underlying APIs (Management API vs PostgREST). Includes mock system for testing without external dependencies.

vs alternatives

Reduces duplication across multiple MCP server packages vs monolithic single-server approach, while providing testing utilities that eliminate need for live API credentials during development

oauth 2.1 dynamic client registration for secure credential exchange

Medium confidence

Implements OAuth 2.1 Dynamic Client Registration protocol for HTTP transport, enabling secure credential exchange without storing long-lived API keys. When an MCP client connects via HTTP, the server initiates OAuth registration to obtain temporary access tokens scoped to specific operations. This eliminates the need to embed Supabase API keys in client configuration files, improving security for multi-user deployments. The implementation handles token refresh and credential lifecycle management.

Solves for

Connect to Supabase MCP server securely without storing API keys in client configEnable multiple users to access the same MCP server with different Supabase projectsImplement secure credential exchange for shared MCP server deploymentsRevoke access to Supabase MCP without rotating API keys

Best for

Teams deploying shared Supabase MCP servers for multiple users

Organizations with strict credential management policies

Multi-tenant MCP deployments requiring per-user authentication

Requires

Supabase account with OAuth 2.1 support

OAuth 2.1 provider configuration (client ID, client secret, redirect URI)

HTTP transport for MCP server (not available for stdio or SSE)

Limitations

OAuth 2.1 setup requires Supabase account configuration — adds operational complexity

Token refresh adds latency to requests — ~100-200ms per token refresh cycle

Dynamic Client Registration is not supported by all OAuth providers — may require custom implementation

What makes it unique

Uses OAuth 2.1 Dynamic Client Registration instead of static API keys, enabling credential-less client configuration and per-session token management. Handles token lifecycle (issuance, refresh, revocation) transparently to clients.

vs alternatives

More secure than API key-based authentication because tokens are temporary and scoped, and more flexible than static credentials because multiple users can access the same server with different permissions

monorepo workspace management with pnpm and biome tooling

Medium confidence

Organizes the Supabase MCP project as a TypeScript monorepo using pnpm workspaces for dependency management and Biome for code formatting and linting. The monorepo structure enables shared utilities (@supabase/mcp-utils) to be used by multiple server implementations (@supabase/mcp-server-supabase, @supabase/mcp-server-postgrest) while maintaining independent versioning and publishing. Biome provides fast, unified linting and formatting across all packages without external tool dependencies.

Solves for

Develop multiple related MCP server packages with shared utilitiesMaintain consistent code style and quality across monorepo packagesPublish independent npm packages from a single repositoryManage dependencies efficiently across multiple packages

Best for

Teams maintaining multiple related npm packages

Projects requiring shared utilities across multiple implementations

Organizations standardizing on pnpm and Biome tooling

Requires

pnpm 8.0+ for workspace support

Node.js 18+

Understanding of monorepo patterns and workspace dependencies

Limitations

pnpm workspace setup requires understanding of workspace protocols — steeper learning curve than npm

Biome is relatively new — may lack some linting rules compared to ESLint

Monorepo adds complexity for simple projects — overhead not justified for single-package repositories

What makes it unique

Uses pnpm workspaces with explicit workspace protocol for dependency management, enabling efficient sharing of utilities while maintaining independent package versioning. Biome provides unified linting/formatting without external tool dependencies.

vs alternatives

More efficient than npm workspaces because pnpm uses symlinks and hoisting, and faster than ESLint+Prettier because Biome is a single unified tool written in Rust

mock system for testing mcp servers without live api access

Medium confidence

Provides a testing framework and mock API system that allows MCP server implementations to be tested without requiring live Supabase API credentials or network access. The mock system (@supabase/mcp-utils) includes predefined mock responses for common operations and allows custom mock configurations for testing edge cases. Tests can verify tool behavior, error handling, and response formatting without external dependencies.

Solves for

Test MCP server implementations in CI/CD pipelines without API credentialsVerify tool behavior and error handling without live API accessDevelop MCP server features locally without requiring Supabase projectTest edge cases and error scenarios that are difficult to reproduce with live APIs

Best for

MCP server developers building Supabase integrations

CI/CD pipelines requiring credential-free testing

Teams practicing test-driven development for MCP servers

Requires

@supabase/mcp-utils package with mock system

TypeScript test framework (Jest, Vitest, etc.)

Understanding of mock configuration patterns

Limitations

Mock responses are predefined — custom scenarios require code changes

Mock system may not cover all API edge cases — gaps between mock and real API behavior

Testing with mocks cannot catch API breaking changes — requires integration tests with live API

What makes it unique

Provides mock system integrated into @supabase/mcp-utils, allowing all MCP server implementations to share testing infrastructure. Mocks are TypeScript-based, enabling type-safe test configurations.

vs alternatives

Eliminates need for live API credentials in tests vs integration testing, while providing more realistic behavior than simple stub responses because mocks include error scenarios and edge cases

supabase management api operations for project administration

Medium confidence

Exposes Supabase Management API endpoints through MCP tools for administrative operations including account management, project configuration, database branching, and cost tracking. The @supabase/mcp-server-supabase package implements a platform abstraction layer that translates MCP tool invocations into authenticated Management API HTTP requests, handling OAuth 2.1 authentication via Dynamic Client Registration and managing API response parsing. Supports feature groups that can be selectively enabled/disabled to control which operations are available.

Solves for

Create, list, and manage Supabase projects from an AI assistantInvite team members and manage project access through an LLM interfaceTrigger database branching and preview deployments for development workflowsMonitor project costs and usage metrics through AI-driven analysis

Best for

DevOps engineers automating Supabase project provisioning

Development teams using AI assistants for infrastructure management

Organizations building internal tools for Supabase administration

Requires

Supabase account with Management API access

OAuth 2.1 credentials or API key for Management API authentication

Appropriate IAM permissions for requested operations (project creation, user management, etc.)

Limitations

Requires Management API credentials with appropriate scopes — not all operations available to all users

Feature groups must be explicitly enabled — some operations disabled by default for safety

Rate limiting on Management API applies — bulk operations may require throttling

What makes it unique

Implements feature groups (selectively enabled tool categories) and platform abstraction that allows the same MCP server to support multiple Supabase deployment modes (cloud, local CLI, self-hosted) with different API endpoints. Uses OAuth 2.1 Dynamic Client Registration for secure credential handling without storing long-lived tokens.

vs alternatives

Provides unified MCP interface to Management API vs scattered CLI commands and web UI, enabling programmatic infrastructure automation while maintaining security through OAuth and feature group controls

database branching and preview deployment orchestration

Medium confidence

Manages Supabase database branching workflows through MCP tools that create, list, and manage preview deployments. The implementation coordinates with the Management API to provision isolated database branches for development/testing, track branch metadata, and trigger deployments. Supports branching workflows where an LLM can create a branch, apply migrations, run tests, and merge changes back to production through a sequence of MCP tool invocations.

Solves for

Create a database branch for feature development without manual CLI commandsLet an AI assistant manage the full lifecycle of a preview deploymentAutomate database schema migrations and testing across branches

Best for

Development teams using AI-assisted database schema management

CI/CD pipelines that need programmatic branch management

Teams practicing database-driven development with preview environments

Requires

Supabase Pro or higher plan with branching feature enabled

Management API credentials with branching permissions

Database migration tools (Supabase CLI or similar) for schema changes

Limitations

Branch creation is asynchronous — requires polling or webhook integration to track completion

Preview deployments have limited lifetime — automatic cleanup after inactivity period

Schema synchronization between branches requires explicit migration tools — not automatic

What makes it unique

Integrates branching operations as part of the broader MCP tool ecosystem, allowing LLMs to orchestrate multi-step workflows (create branch → migrate schema → test → merge) through sequential tool calls. Abstracts async operations and polling complexity from the LLM caller.

vs alternatives

Enables programmatic branching workflows vs manual CLI usage, integrating with LLM reasoning to automate complex database development scenarios

authentication user and session management through auth api

Medium confidence

Provides MCP tools for managing Supabase Auth users, sessions, and authentication policies. Integrates with Supabase Auth API to create users, reset passwords, manage MFA, list active sessions, and configure authentication settings. The implementation handles Auth API authentication and translates MCP tool parameters into Auth API requests, returning user objects and session metadata in structured JSON format.

Solves for

Create and manage user accounts in my Supabase Auth system from an AI assistantReset user passwords or manage MFA settings programmaticallyList active sessions and revoke compromised authentication tokensConfigure authentication policies (password requirements, session timeouts) through an LLM interface

Best for

Support teams using AI assistants for user account management

Developers building admin dashboards with AI-assisted user operations

Security teams automating authentication policy enforcement

Requires

Supabase project with Auth enabled

Auth API credentials with appropriate scopes (user management, session management)

Valid user identifiers (email, user ID, or phone number depending on operation)

Limitations

Auth API operations are subject to rate limiting — bulk user creation requires throttling

Password reset tokens have limited lifetime — must be used immediately

MFA management requires user cooperation — cannot force MFA enrollment through API alone

What makes it unique

Separates Auth operations into dedicated MCP tools with explicit safety considerations (rate limiting awareness, token lifetime management). Integrates with Supabase Auth's built-in security features (password hashing, MFA) rather than reimplementing them.

vs alternatives

Provides safe, rate-limited user management through MCP vs direct Auth API access, with built-in error handling for common auth scenarios (invalid tokens, rate limits, user conflicts)

cloud storage bucket and object management with signed urls

Medium confidence

Manages Supabase Storage buckets and objects through MCP tools that support uploading, downloading, listing, and deleting files. The implementation generates signed URLs for secure temporary access, handles multipart uploads for large files, and manages bucket policies and access controls. Integrates with Supabase Storage API to perform operations and returns file metadata, signed URLs, and operation status in structured format.

Solves for

Upload files to Supabase Storage from an AI assistant workflowGenerate temporary signed URLs for secure file sharingList and manage files in storage buckets programmaticallyDelete or organize files in storage through an LLM interface

Best for

Developers building file management features with AI assistance

Content teams using AI assistants to organize media libraries

Applications that need programmatic storage access without exposing credentials

Requires

Supabase project with Storage enabled

Storage API credentials with appropriate bucket permissions

Bucket names and object paths for operations

Limitations

Signed URL generation requires valid Storage API credentials — cannot be done client-side without key exposure

Large file uploads may timeout — multipart upload support required for files >100MB

Bucket policies are immutable after creation — cannot change bucket privacy settings through API

What makes it unique

Generates signed URLs through MCP tools, allowing LLMs to create temporary access tokens without exposing long-lived credentials. Handles multipart upload complexity internally, presenting simple upload interface to callers.

vs alternatives

Safer than exposing Storage API keys directly because it generates time-limited signed URLs, and more flexible than static file serving because it supports dynamic access control and programmatic file organization

edge function invocation and deployment management

Medium confidence

Invokes Supabase Edge Functions and manages their deployment through MCP tools. The implementation handles function discovery, parameter passing, and response parsing for both synchronous and asynchronous function invocations. Supports deploying new functions, updating existing functions, and managing function environment variables and secrets through Management API integration.

Solves for

Invoke a Supabase Edge Function from an AI assistant to perform backend logicDeploy or update Edge Functions programmatically without manual CLI commandsManage function environment variables and secrets through an LLM interfaceLet Claude or Cursor trigger serverless functions as part of automated workflows

Best for

Developers building serverless workflows with AI orchestration

Teams automating Edge Function deployment and updates

Applications that need dynamic function invocation based on LLM reasoning

Requires

Supabase project with Edge Functions enabled

Edge Function names and parameter specifications

Management API credentials for deployment operations

Limitations

Function invocation timeout is 600 seconds — long-running operations must be async

Function response size limited to ~6MB — large payloads require streaming or storage

Environment variables and secrets require Management API credentials — not available through PostgREST

What makes it unique

Separates function invocation (runtime) from function deployment (management), allowing LLMs to both trigger functions and manage their lifecycle. Handles async deployment operations transparently through polling abstraction.

vs alternatives

Enables programmatic function invocation and deployment vs manual CLI usage, allowing LLMs to orchestrate serverless workflows and manage function configuration as code

multi-transport mcp server with http, stdio, and sse support

Medium confidence

Implements the Model Context Protocol across multiple transport mechanisms (HTTP with OAuth, stdio for local CLI, Server-Sent Events for streaming) through a unified server implementation. The core server creation logic (@supabase/mcp-server-supabase) abstracts transport details, allowing the same tool definitions and business logic to be exposed through different channels. HTTP transport includes OAuth 2.1 Dynamic Client Registration for secure credential exchange, while stdio and SSE transports support local development and streaming scenarios.

Solves for

Connect to Supabase MCP server through my preferred transport (HTTP, stdio, or SSE)Use Supabase MCP with Claude Desktop (HTTP), local Supabase CLI (stdio), or streaming clients (SSE)Deploy Supabase MCP server in different environments (cloud, local, self-hosted) with appropriate transport

Best for

MCP client developers supporting multiple transport mechanisms

Teams deploying Supabase MCP in diverse environments (cloud, local, self-hosted)

Developers building custom MCP clients with specific transport requirements

Requires

MCP-compatible client supporting at least one transport mechanism

For HTTP: OAuth 2.1 provider configuration and Dynamic Client Registration support

For stdio: Node.js 18+ and direct process spawning capability

Limitations

HTTP transport requires OAuth 2.1 setup — adds authentication complexity vs simple API keys

Stdio transport is single-connection — not suitable for concurrent client access

SSE transport is unidirectional (server-to-client) — requires separate channel for client-to-server communication

What makes it unique

Uses a platform abstraction layer that decouples tool definitions from transport implementation, enabling single codebase to support HTTP (with OAuth), stdio (for local CLI), and SSE (for streaming) without duplicating business logic. Transport selection is configurable at server initialization.

vs alternatives

More flexible than single-transport implementations because it supports diverse deployment scenarios (cloud HTTP, local CLI, self-hosted), while maintaining consistent tool definitions across all transports

cost tracking and usage analytics through management api

Medium confidence

Exposes Supabase project cost and usage metrics through MCP tools that query the Management API for billing information, resource consumption, and cost projections. The implementation retrieves metrics for database operations, storage usage, function invocations, and other billable resources, presenting them in structured format for analysis and reporting. Supports filtering by time period and resource type.

Solves for

Monitor my Supabase project costs and usage from an AI assistantGet cost projections and usage alerts through an LLM interfaceAnalyze resource consumption patterns to optimize spendingGenerate cost reports for billing and budgeting purposes

Best for

Finance teams tracking cloud infrastructure costs

DevOps engineers optimizing Supabase resource usage

Developers building cost monitoring dashboards with AI assistance

Requires

Supabase project with billing enabled

Management API credentials with billing/cost access permissions

Project identifier for cost queries

Limitations

Cost data is updated with delay (typically 24 hours) — not real-time

Usage metrics are aggregated by day — sub-daily granularity not available

Cost projections are estimates based on historical usage — actual costs may vary

What makes it unique

Integrates cost and usage data from Management API into MCP tools, enabling LLMs to analyze spending patterns and make optimization recommendations. Handles API rate limiting and data aggregation transparently.

vs alternatives

Provides programmatic cost analysis vs manual dashboard inspection, enabling AI-driven cost optimization and automated alerting based on usage thresholds

feature group-based capability control and safety gating

Medium confidence

Implements selective tool availability through feature groups that can be enabled/disabled at server initialization. The server configuration allows operators to control which MCP tools are exposed (e.g., enable database operations but disable destructive Management API operations), providing safety gating for different deployment scenarios. Feature groups are defined in server configuration and enforced at tool registration time, preventing unauthorized operations without requiring runtime checks.

Solves for

Restrict which Supabase operations are available in a given MCP server instanceEnable read-only database access while disabling destructive operationsSafely expose Supabase MCP to untrusted clients by limiting available toolsConfigure different feature sets for development vs production deployments

Best for

Teams deploying Supabase MCP to untrusted or semi-trusted clients

Organizations with strict security policies requiring operation whitelisting

Development teams wanting different capability sets for dev/staging/prod environments

Requires

Server configuration file with feature group definitions

Understanding of which tools belong to which feature groups

Server restart required to change feature group settings

Limitations

Feature groups are static per server instance — cannot be changed at runtime

No granular permission control within a feature group — all-or-nothing per group

Feature group configuration is server-side only — clients cannot request specific capabilities

What makes it unique

Implements feature groups as first-class configuration concept rather than runtime permission checks, enabling static analysis of available capabilities and preventing accidental exposure of dangerous operations. Feature groups are enforced at tool registration time, not invocation time.

vs alternatives

Safer than runtime permission checks because dangerous operations are not registered at all, and more flexible than hardcoded tool lists because feature groups can be configured per deployment

Capabilities are decomposed by AI analysis. Each maps to specific user intents and improves with match feedback.

Related Artifactssharing capabilities

Artifacts that share capabilities with Supabase MCP Server, ranked by overlap. Discovered automatically through the match graph.

MCP Server20

@supabase/mcp-server-postgrest

MCP server for PostgREST

postgrest api schema introspection and tool registrationparameterized database query execution via mcp tool callsmcp server lifecycle and tool registry management

3 shared capabilities

MCP Server24

user-postgresql-mcp

A PostgreSQL MCP server built with @modelcontextprotocol/sdk.

mcp tool schema registration for database operationspostgresql query execution via mcp protocolmcp protocol message handling and routing

3 shared capabilities

MCP Server37

@supabase/mcp-server-supabase

MCP server for interacting with Supabase

sql query execution with supabase postgresql backendtable schema introspection and metadata retrievaldatabase function invocation with postgresql stored procedures

3 shared capabilities

MCP Server23

Supabase

** - Connects to Supabase platform for database, auth, edge functions and more.

database schema introspection and postgrest api exposuremcp-standardized supabase management api tool exposure

2 shared capabilities

MCP Server34

@supabase/mcp-server-supabase

MCP server for interacting with Supabase

postgresql table crud operations via mcp protocolschema introspection and capability discovery

2 shared capabilities

MCP Server28

@iflow-mcp/garethcott_enhanced-postgres-mcp-server

Enhanced PostgreSQL MCP server with read and write capabilities. Based on @modelcontextprotocol/server-postgres by Anthropic.

mcp resource-based database schema introspectionpostgresql query execution via mcp protocol

2 shared capabilities

Best For

✓AI assistant users (Claude, Cursor, Windsurf) wanting native Supabase integration
✓Teams building LLM agents that need backend-as-a-service capabilities
✓Developers migrating from custom REST integrations to standardized MCP protocol
✓Developers building LLM agents that need read/write database access
✓Data analysts using AI assistants to explore and transform Supabase data
✓Teams automating database operations through AI-driven workflows
✓MCP server developers building Supabase integrations
✓Teams extending Supabase MCP with custom tools

Known Limitations

⚠Requires MCP-compatible client — not usable with standard REST API clients
⚠Tool discovery is static per server initialization — dynamic tool registration not supported
⚠Schema generation depends on Supabase API stability — breaking API changes require server updates
⚠PostgREST API limitations apply — complex joins and aggregations may require raw SQL
⚠Row-level security (RLS) policies are enforced — LLM cannot bypass database-level access controls
⚠Large result sets may exceed MCP message size limits — pagination required for bulk queries

Requirements

MCP-compatible client (Claude Desktop, Cursor, VS Code Copilot, or Windsurf)Supabase project with valid API credentials (project URL and API key)Network access to Supabase Management API or local Supabase instanceSupabase project with PostgREST API enabledValid database credentials with appropriate table-level permissions@supabase/mcp-server-postgrest package installed and configuredNetwork access to PostgREST endpoint (typically https://<project>.supabase.co/rest/v1)TypeScript 5.0+ and Node.js 18+

Input / Output

Accepts: JSON-formatted tool arguments matching MCP schema definitions, Structured parameters for database queries, auth operations, storage access, SQL query strings with parameter placeholders, Table names and column specifications for CRUD operations, Filter conditions and sorting parameters, Tool schema definitions (JSON Schema format), Platform interface implementations, Mock API response configurations, OAuth 2.1 authorization codes from client, Token refresh requests, TypeScript source files, package.json configurations, Biome configuration (biome.json), Mock response configurations, Test case definitions, Tool invocation parameters, Project identifiers and configuration parameters, User email addresses and role specifications, Database branch names and source specifications, Branch names and source branch identifiers, Branch configuration parameters (compute size, retention period), Migration SQL or schema definitions, User email addresses and metadata, Password reset tokens and new password values, MFA configuration parameters, Session identifiers for revocation, File content (binary or base64-encoded), Bucket names and object paths, Signed URL expiration times, Bucket policy configurations, Function names and invocation parameters, Function source code for deployment, Environment variable and secret key-value pairs, MCP protocol messages (JSON-RPC format), Tool invocation requests with parameters, Resource requests and capability discovery queries, Time period specifications (start date, end date), Resource type filters (database, storage, functions, etc.), Feature group names (e.g., 'database', 'auth', 'storage', 'management'), Boolean enable/disable flags per feature group

Produces: JSON-structured tool results, Query result sets from database operations, Status responses from auth and storage operations, JSON arrays of query result rows, Metadata objects with row counts and operation status, Error messages with database constraint violation details, Reusable tool builders and schema validators, Platform abstraction implementations, Mock API responses for testing, OAuth access tokens with expiration, Token refresh responses, Authorization error messages, Formatted and linted code, Build artifacts for each package, Published npm packages, Mock API responses, Test results and assertions, Coverage reports, Project metadata and configuration objects, User and team membership lists, Branch creation status and deployment information, Cost and usage metrics in structured format, Branch metadata (ID, status, connection string), Deployment status and progress information, Migration execution results and error logs, User objects with profile metadata, Session lists with creation/expiration timestamps, Authentication policy configuration objects, Operation status and error messages, File metadata (size, content-type, created-at), Signed URLs with expiration timestamps, Bucket listings with object summaries, Function invocation results (JSON or text), Function metadata and configuration objects, Error messages and execution logs, MCP protocol responses (JSON-RPC format), Tool results and error messages, Resource content and capability definitions, Cost summaries with breakdown by resource type, Usage metrics (GB stored, function invocations, etc.), Cost projections and trend analysis, Billing period information, Tool availability information in MCP capability discovery, Error messages when disabled tools are invoked

UnfragileRank

Adoption70%(30% weight)

Quality23%(25% weight)

Ecosystem52%(25% weight)

Match Graph10%(15% weight)

Freshness100%(5% weight)

UnfragileRank is computed from adoption signals, documentation quality, ecosystem connectivity, match graph feedback, and freshness. No artifact can pay for a higher rank.

Type: MCP Server

14 capabilities

Visit Supabase MCP Server→

About

Community MCP server for Supabase backend-as-a-service. Provides tools for database queries, table management, auth user operations, storage bucket access, and edge function invocation.

Alternatives to Supabase MCP Server

YouTube MCP Server46MCP Server

Extract and analyze YouTube video transcripts via MCP.

Compare →

Vercel MCP Server46MCP Server

Manage Vercel deployments, projects, and domains via MCP.

Compare →

Todoist MCP Server46MCP Server

Create and manage Todoist tasks and projects via MCP.

Compare →

Telegram MCP Server46MCP Server

Send messages and manage Telegram chats and bots via MCP.

Compare →

Are you the builder of Supabase MCP Server?

Claim this artifact to get a verified badge, access match analytics, see which intents users search for, and manage your listing.

Claim this artifact →Verification via email

Get the weekly brief

New tools, rising stars, and what's actually worth your time. No spam.

Data Sources

seed developer essentials

Looking for something else?

Search →

Capabilities14 decomposed

mcp-standardized supabase api tool registration and schema exposure

Medium confidence

Solves for

Best for

AI assistant users (Claude, Cursor, Windsurf) wanting native Supabase integration

Teams building LLM agents that need backend-as-a-service capabilities

Developers migrating from custom REST integrations to standardized MCP protocol

Requires

MCP-compatible client (Claude Desktop, Cursor, VS Code Copilot, or Windsurf)

Supabase project with valid API credentials (project URL and API key)

Network access to Supabase Management API or local Supabase instance

Limitations

Requires MCP-compatible client — not usable with standard REST API clients

Tool discovery is static per server initialization — dynamic tool registration not supported

Schema generation depends on Supabase API stability — breaking API changes require server updates

What makes it unique

vs alternatives

Provides protocol-standard tool exposure vs proprietary integrations, enabling use across any MCP-compatible client without vendor lock-in or duplicate integration work

postgresql database query execution with postgrest api abstraction

Medium confidence

Solves for

Best for

Developers building LLM agents that need read/write database access

Data analysts using AI assistants to explore and transform Supabase data

Teams automating database operations through AI-driven workflows

Requires

Supabase project with PostgREST API enabled

Valid database credentials with appropriate table-level permissions

@supabase/mcp-server-postgrest package installed and configured

Limitations

PostgREST API limitations apply — complex joins and aggregations may require raw SQL

Row-level security (RLS) policies are enforced — LLM cannot bypass database-level access controls

Large result sets may exceed MCP message size limits — pagination required for bulk queries

What makes it unique

vs alternatives

shared utility abstractions for mcp server development

Medium confidence

Solves for

Best for

MCP server developers building Supabase integrations

Teams extending Supabase MCP with custom tools

Contributors to the Supabase MCP project

Requires

TypeScript 5.0+ and Node.js 18+

@supabase/mcp-utils package installed

Understanding of MCP protocol and Supabase API patterns

Limitations

Utilities are TypeScript-only — no Python or other language support

Mock system is limited to predefined scenarios — custom mocking requires code changes

Abstractions add complexity for simple use cases — overhead not justified for single-tool servers

What makes it unique

vs alternatives

Reduces duplication across multiple MCP server packages vs monolithic single-server approach, while providing testing utilities that eliminate need for live API credentials during development

oauth 2.1 dynamic client registration for secure credential exchange

Medium confidence

Solves for

Best for

Teams deploying shared Supabase MCP servers for multiple users

Organizations with strict credential management policies

Multi-tenant MCP deployments requiring per-user authentication

Requires

Supabase account with OAuth 2.1 support

OAuth 2.1 provider configuration (client ID, client secret, redirect URI)

HTTP transport for MCP server (not available for stdio or SSE)

Limitations

OAuth 2.1 setup requires Supabase account configuration — adds operational complexity

Token refresh adds latency to requests — ~100-200ms per token refresh cycle

Dynamic Client Registration is not supported by all OAuth providers — may require custom implementation

What makes it unique

vs alternatives

monorepo workspace management with pnpm and biome tooling

Medium confidence

Solves for

Best for

Teams maintaining multiple related npm packages

Projects requiring shared utilities across multiple implementations

Organizations standardizing on pnpm and Biome tooling

Requires

pnpm 8.0+ for workspace support

Node.js 18+

Understanding of monorepo patterns and workspace dependencies

Limitations

pnpm workspace setup requires understanding of workspace protocols — steeper learning curve than npm

Biome is relatively new — may lack some linting rules compared to ESLint

Monorepo adds complexity for simple projects — overhead not justified for single-package repositories

What makes it unique

vs alternatives

More efficient than npm workspaces because pnpm uses symlinks and hoisting, and faster than ESLint+Prettier because Biome is a single unified tool written in Rust

mock system for testing mcp servers without live api access

Medium confidence

Solves for

Best for

MCP server developers building Supabase integrations

CI/CD pipelines requiring credential-free testing

Teams practicing test-driven development for MCP servers

Requires

@supabase/mcp-utils package with mock system

TypeScript test framework (Jest, Vitest, etc.)

Understanding of mock configuration patterns

Limitations

Mock responses are predefined — custom scenarios require code changes

Mock system may not cover all API edge cases — gaps between mock and real API behavior

Testing with mocks cannot catch API breaking changes — requires integration tests with live API

What makes it unique

Provides mock system integrated into @supabase/mcp-utils, allowing all MCP server implementations to share testing infrastructure. Mocks are TypeScript-based, enabling type-safe test configurations.

vs alternatives

Eliminates need for live API credentials in tests vs integration testing, while providing more realistic behavior than simple stub responses because mocks include error scenarios and edge cases

supabase management api operations for project administration

Medium confidence

Solves for

Best for

DevOps engineers automating Supabase project provisioning

Development teams using AI assistants for infrastructure management

Organizations building internal tools for Supabase administration

Requires

Supabase account with Management API access

OAuth 2.1 credentials or API key for Management API authentication

Appropriate IAM permissions for requested operations (project creation, user management, etc.)

Limitations

Requires Management API credentials with appropriate scopes — not all operations available to all users

Feature groups must be explicitly enabled — some operations disabled by default for safety

Rate limiting on Management API applies — bulk operations may require throttling

What makes it unique

vs alternatives

database branching and preview deployment orchestration

Medium confidence

Solves for

Best for

Development teams using AI-assisted database schema management

CI/CD pipelines that need programmatic branch management

Teams practicing database-driven development with preview environments

Requires

Supabase Pro or higher plan with branching feature enabled

Management API credentials with branching permissions

Database migration tools (Supabase CLI or similar) for schema changes

Limitations

Branch creation is asynchronous — requires polling or webhook integration to track completion

Preview deployments have limited lifetime — automatic cleanup after inactivity period

Schema synchronization between branches requires explicit migration tools — not automatic

What makes it unique

vs alternatives

Enables programmatic branching workflows vs manual CLI usage, integrating with LLM reasoning to automate complex database development scenarios

authentication user and session management through auth api

Medium confidence

Solves for

Best for

Support teams using AI assistants for user account management

Developers building admin dashboards with AI-assisted user operations

Security teams automating authentication policy enforcement

Requires

Supabase project with Auth enabled

Auth API credentials with appropriate scopes (user management, session management)

Valid user identifiers (email, user ID, or phone number depending on operation)

Limitations

Auth API operations are subject to rate limiting — bulk user creation requires throttling

Password reset tokens have limited lifetime — must be used immediately

MFA management requires user cooperation — cannot force MFA enrollment through API alone

What makes it unique

vs alternatives

Provides safe, rate-limited user management through MCP vs direct Auth API access, with built-in error handling for common auth scenarios (invalid tokens, rate limits, user conflicts)

cloud storage bucket and object management with signed urls

Medium confidence

Solves for

Best for

Developers building file management features with AI assistance

Content teams using AI assistants to organize media libraries

Applications that need programmatic storage access without exposing credentials

Requires

Supabase project with Storage enabled

Storage API credentials with appropriate bucket permissions

Bucket names and object paths for operations

Limitations

Signed URL generation requires valid Storage API credentials — cannot be done client-side without key exposure

Large file uploads may timeout — multipart upload support required for files >100MB

Bucket policies are immutable after creation — cannot change bucket privacy settings through API

What makes it unique

vs alternatives

edge function invocation and deployment management

Medium confidence

Solves for

Best for

Developers building serverless workflows with AI orchestration

Teams automating Edge Function deployment and updates

Applications that need dynamic function invocation based on LLM reasoning

Requires

Supabase project with Edge Functions enabled

Edge Function names and parameter specifications

Management API credentials for deployment operations

Limitations

Function invocation timeout is 600 seconds — long-running operations must be async

Function response size limited to ~6MB — large payloads require streaming or storage

Environment variables and secrets require Management API credentials — not available through PostgREST

What makes it unique

vs alternatives

Enables programmatic function invocation and deployment vs manual CLI usage, allowing LLMs to orchestrate serverless workflows and manage function configuration as code

multi-transport mcp server with http, stdio, and sse support

Medium confidence

Solves for

Best for

MCP client developers supporting multiple transport mechanisms

Teams deploying Supabase MCP in diverse environments (cloud, local, self-hosted)

Developers building custom MCP clients with specific transport requirements

Requires

MCP-compatible client supporting at least one transport mechanism

For HTTP: OAuth 2.1 provider configuration and Dynamic Client Registration support

For stdio: Node.js 18+ and direct process spawning capability

Limitations

HTTP transport requires OAuth 2.1 setup — adds authentication complexity vs simple API keys

Stdio transport is single-connection — not suitable for concurrent client access

SSE transport is unidirectional (server-to-client) — requires separate channel for client-to-server communication

What makes it unique

vs alternatives

cost tracking and usage analytics through management api

Medium confidence

Solves for

Best for

Finance teams tracking cloud infrastructure costs

DevOps engineers optimizing Supabase resource usage

Developers building cost monitoring dashboards with AI assistance

Requires

Supabase project with billing enabled

Management API credentials with billing/cost access permissions

Project identifier for cost queries

Limitations

Cost data is updated with delay (typically 24 hours) — not real-time

Usage metrics are aggregated by day — sub-daily granularity not available

Cost projections are estimates based on historical usage — actual costs may vary

What makes it unique

vs alternatives

Provides programmatic cost analysis vs manual dashboard inspection, enabling AI-driven cost optimization and automated alerting based on usage thresholds

feature group-based capability control and safety gating

Medium confidence

Solves for

Best for

Teams deploying Supabase MCP to untrusted or semi-trusted clients

Organizations with strict security policies requiring operation whitelisting

Development teams wanting different capability sets for dev/staging/prod environments

Requires

Server configuration file with feature group definitions

Understanding of which tools belong to which feature groups

Server restart required to change feature group settings

Limitations

Feature groups are static per server instance — cannot be changed at runtime

No granular permission control within a feature group — all-or-nothing per group

Feature group configuration is server-side only — clients cannot request specific capabilities

What makes it unique

vs alternatives

Safer than runtime permission checks because dangerous operations are not registered at all, and more flexible than hardcoded tool lists because feature groups can be configured per deployment

Capabilities are decomposed by AI analysis. Each maps to specific user intents and improves with match feedback.

Alternatives to Supabase MCP Server

YouTube MCP Server46MCP Server

Extract and analyze YouTube video transcripts via MCP.

Compare →

Vercel MCP Server46MCP Server

Manage Vercel deployments, projects, and domains via MCP.

Compare →

Todoist MCP Server46MCP Server

Create and manage Todoist tasks and projects via MCP.

Compare →

Telegram MCP Server46MCP Server

Send messages and manage Telegram chats and bots via MCP.

Compare →

Supabase MCP Server

Capabilities14 decomposed

mcp-standardized supabase api tool registration and schema exposure

postgresql database query execution with postgrest api abstraction

shared utility abstractions for mcp server development

oauth 2.1 dynamic client registration for secure credential exchange

monorepo workspace management with pnpm and biome tooling

mock system for testing mcp servers without live api access

supabase management api operations for project administration

database branching and preview deployment orchestration

authentication user and session management through auth api

cloud storage bucket and object management with signed urls

edge function invocation and deployment management

multi-transport mcp server with http, stdio, and sse support

cost tracking and usage analytics through management api

feature group-based capability control and safety gating

Related Artifactssharing capabilities

@supabase/mcp-server-postgrest

user-postgresql-mcp

@supabase/mcp-server-supabase

Supabase

@supabase/mcp-server-supabase

@iflow-mcp/garethcott_enhanced-postgres-mcp-server

Best For

Known Limitations

Requirements

Input / Output

UnfragileRank

About

Categories

Alternatives to Supabase MCP Server

Are you the builder of Supabase MCP Server?

Get the weekly brief

Data Sources

Supabase MCP Server

Capabilities14 decomposed

mcp-standardized supabase api tool registration and schema exposure

postgresql database query execution with postgrest api abstraction

shared utility abstractions for mcp server development

oauth 2.1 dynamic client registration for secure credential exchange

monorepo workspace management with pnpm and biome tooling

mock system for testing mcp servers without live api access

supabase management api operations for project administration

database branching and preview deployment orchestration

authentication user and session management through auth api

cloud storage bucket and object management with signed urls

edge function invocation and deployment management

multi-transport mcp server with http, stdio, and sse support

cost tracking and usage analytics through management api

feature group-based capability control and safety gating

Related Artifactssharing capabilities

@supabase/mcp-server-postgrest

user-postgresql-mcp

@supabase/mcp-server-supabase

Supabase

@supabase/mcp-server-supabase

@iflow-mcp/garethcott_enhanced-postgres-mcp-server

Best For

Known Limitations

Requirements

Input / Output

UnfragileRank

About

Categories

Alternatives to Supabase MCP Server

Are you the builder of Supabase MCP Server?

Get the weekly brief

Data Sources