| Product | Monospace |
| Website | monospace.io |
| Category | Data Infrastructure / API Layer |
| Built By | Directus team |
| GA Date | October 1, 2026 |
| Protocols | REST, GraphQL, MCP |
| Access Control | Row and field-level permissions |
What Is Monospace?
Monospace is a governed API layer that sits between your data sources and everyone who builds on them. It connects databases, APIs, and SaaS systems and gives developers, business teams, and AI agents live read-write access under one permissions model — without copying or moving data.
It comes from the team behind Directus, the open-source data platform. Where Directus wraps a single database, Monospace federates across multiple sources: your Postgres database, your Stripe account, your Salesforce instance, your internal APIs. One query can join data from all of them.
Key Features
Federated Queries
Query across databases, APIs, and SaaS tools in a single request. No ETL pipelines, no data warehouses, no stale copies. Monospace resolves the query against each connected source in real time and returns a unified result. This eliminates the “data is in three places and none of them agree” problem.
Granular Permissions
Row-level and field-level permissions per caller. A marketing team member sees campaign data but not salary data. An AI agent reads product catalog but can’t write financial records. Permissions are centralized in Monospace rather than configured separately in every data source.
MCP Integration
Monospace exposes data through MCP (Model Context Protocol), so AI agents like Claude, ChatGPT, and Cursor can access governed data directly. The permissions model applies equally to AI agents and human users — an agent can only see and modify what its permission set allows.
REST and GraphQL
Traditional API consumers get both REST and GraphQL endpoints. The same permission model governs all three protocols (REST, GraphQL, MCP), so adding a new consumer doesn’t mean rewriting access controls.
Who Is This For?
- Teams with data scattered across multiple systems who need a single API layer
- Companies deploying AI agents that need governed access to enterprise data
- Platform teams building internal APIs for multiple consumer teams
- Organizations with strict data governance requirements (healthcare, finance, legal)
Pros
- Federates across databases, APIs, and SaaS — no data copying
- Row and field-level permissions in one place
- MCP integration for governed AI agent access
- REST + GraphQL + MCP from one layer
- Directus team has proven data platform experience
- Live read-write, not just read-only
Cons
- Just reached GA — expect rough edges
- Federated queries add latency vs direct access
- Pricing not yet published
- Competes with established iPaaS and data mesh tools
- Permission model complexity increases with each connected source
Verdict
Monospace addresses a real problem at the right time. Every company with AI agents needs governed data access, and most are solving it with custom middleware, API gateways, and per-source permission configs. Monospace consolidates this into a single layer with a single permission model.
The MCP integration is the timely feature. As AI agents move from prototypes to production systems, the question of “what data can this agent access?” becomes critical. Monospace answers it with the same granular controls that govern human access.
The Directus pedigree is relevant — they’ve been building data access layers for years and have a large open-source community. Monospace is the next evolution of that work, expanded from single-database to multi-source federation.
Build a data platform?
Get reviewed and linked from our 27-site network. Placements live in 48 hours.
See Packages