Skip to main content
These pages are for running Nasiko yourself. The open-source server is the default; an Enterprise license adds the self-managed extras marked on each page. On Nasiko Cloud, Nasiko runs the server for you. Copy .env.example to .env in the repository checkout (nasiko up does this for you if .env is missing), or set the same names in the server’s environment. The server reads them at start. This page lists names. Do not commit .env.

Required to start

Model routing and agents

Observability and coding agents

Flow limits

Also configurable in the dashboard Settings → Flow limits. See Flow limits.

Context budgets and compression

See Reduce cost.

MCP gateway

Seed agents, workflows, and connectors

Seeding sets up shared agents, workflows, and connectors when the server starts, so every user sees them without setting anything up. Everything seeded is owned by the first superuser (the bootstrap admin) and shared with all users. Seeding runs in the background after the server starts, so it never delays startup. An entry that fails is logged and skipped. The server reads these variables once at start; restart it after changing them.

Agent settings

Every field is optional; one you leave out keeps the server’s default. An unknown field or an impossible value stops the server from starting, so a typo can’t silently leave a shared agent under-provisioned.

Connectors

A connector that already exists under that name keeps its stored configuration. Its sharing is re-applied and its tools are re-synced on every start. A bearer or basic entry that also sets a static Authorization header is refused: users without their own credential would silently run on yours. To share one key with everyone, use auth_type: none with the header. Invalid JSON is logged and ignored.

Source-run connection URLs

Used when the server is not the Compose server service:

SSO, GitHub, multi-tenant, imports

Full behavioral reference for flow and hosting: Flow limits, Agent hosting.