In October 2023 AI agents created 0.1% of databases on Neon; by October 2025 they created 80%, and at Fly.io agents have overtaken humans as the most common customer species. Fly.io built mcp.fly.io, an MCP server that connects agents to Sprites for sandboxed development and to Fly Machines for production, providing a path from prototype to scalable deployment. Sprites are real Linux machines with persistent disk, service management, HTTPS URLs, and filesystem checkpoints that allow agents to install packages, run servers, and roll back after failures, and they sleep when idle. When a project needs always-on availability, the agent can migrate it from a Sprite to a Fly Machine using the same MCP server, creating the Machine, volume, and secrets in the same conversation. The MCP server is authorized per app and organization with no API token, and Fly.io is investigating adding MPG skills for database management.
Context
AI agents now create 80% of databases on Neon and have overtaken humans as the most common customer species at Fly.io. This shift demands a cloud infrastructure that supports both sandboxed experimentation and a clear path to production, which motivated Fly.io to build an MCP server tailored for agent workflows.
Approach / What changed
The Fly MCP server (mcp.fly.io) connects AI agents to Fly.io’s Sprites for sandboxed development and to Fly Machines for production. Agents build and test in Sprites, then migrate to Machines for always-on availability, all within the same MCP conversation.
Takeaways
1
Sprites provide a real Linux environment with persistent disk, service management, HTTPS URLs, and filesystem checkpoints, enabling agents to test packages, servers, and restarts realistically.
2
Migration from a Sprite to a Fly Machine is straightforward because both are Linux boxes; the agent built a Linux service and can move it without translating runtimes.
3
The Fly MCP server is authorized per app and organization with no API token, and Fly.io plans to add MPG (Postgres) skills for database management.
FAQ
What is a Sprite?
A Sprite is a real Linux machine with its own kernel, persistent disk, service management, and HTTPS URLs. It can sleep when idle and wake on request, and supports filesystem checkpoints for rollback after failures.
How does migration from a Sprite to a Fly Machine work?
The agent uses the same MCP server to create a Machine, volume, and secrets in the same conversation. Since both Sprites and Machines are Linux boxes running the same service, migration is simply moving a Linux service between two Linux machines without translating runtimes.
Will the Fly MCP server support databases?
Fly.io is investigating adding MPG (Postgres) skills to the MCP server, which would allow the agent to create a Postgres cluster with automatic backups and failover and attach it to the app in the same conversation.
Autonomous Recovery is a closed loop that senses degradation, classifies a fault as transient or hardware, recovers workloads and quarantines hardware in parallel, then verifies against an RTO.
An Agent SDK agent is a folder of instructions, tools, skills, and channels. This walkthrough follows five checkpoints from local validation to gateway-gated tool calls and replayable evals.
Keep reading
Related reading
The Good Stuff
Paid picks we use ourselves. Links carry a referral code — a subscription through one earns us a commission at no extra cost to you.