Floci Studio gives you a full AWS emulator with a beautiful GUI, a Model Context Protocol server your AI agent can drive, and a one-click marketplace for every service your stack depends on. Zero cloud costs. Zero rate limits. Zero context switches.
36 native service views with full CRUD operations. Not a generic JSON explorer — each service has a purpose-built UI: SQS drill-down with ReceiveMessage, SNS FIFO topics, Lambda code editor, S3 object browser, DynamoDB query builder.
The first local AWS emulator built with AI agents in mind. Connect Claude, Cursor, or any MCP-compatible client and let your agent create queues, invoke lambdas, read messages, seed DynamoDB tables, and run end-to-end flows — all via natural language.
One-click recipes for every service your stack depends on: a local AI stack with Ollama + Qdrant, Temporal workflows, NATS JetStream, HashiCorp Vault secrets, ClickHouse analytics, Portainer, and more. Each one maps to a managed AWS service (Postgres→RDS, Redis→ElastiCache, Ollama→Bedrock) so it deploys straight to production — all wired together via Docker Compose.
Coverage
Every service runs against the local floci engine on port 4566 — the same API surface as real AWS, so your code works without modification when you deploy. A green dot marks a purpose-built GUI; the rest are wired through the sidecar and reachable from the UI and your AI agent.
Marketplace
Deploy any supporting service directly from the floci UI. Each recipe is a parameterized Docker Compose template — configure ports, credentials, and persistence, then hit Deploy.
MCP Server
Connect Claude Desktop, Cursor, or any MCP client with a single config block. Your agent gets 119 tools: create queues, publish events, invoke lambdas, seed databases, tag resources across services, generate architecture diagrams, export Terraform — all from a conversation.
Clone the repo, run docker compose up,
open localhost:3000.
That's it.