Deploy and Host Atomic on Railway

About Hosting Atomic

Atomic is a self-hosted personal knowledge base for Markdown notes, semantic connections, search, graph exploration, APIs, WebSockets, and MCP. This template deploys Atomic v1.45.0 from an immutable upstream image.

Common Use Cases

A personal or trusted-team knowledge base.
Structured notes, tags, links, and multi-database organization.
API, browser-extension, mobile, or MCP access with separate tokens.
Optional operator-configured AI-assisted search and synthesis.

Dependencies for Atomic Hosting

Deployment Dependencies

The template creates one public service and one volume at /data. Railway generates a first-claim setup token and sets the public URL reference. The operator claims once, saves the one-time Bearer token, and creates separate tokens per integration.

SQLite registry and knowledge databases persist together under /data. Serverless is disabled and one replica is required. Postgres, Redis, Buckets, Litestream, Ollama, and AI-provider credentials are not included.

Why Deploy Atomic on Railway?

Railway supplies build/image resolution, TLS domain, secret generation, health checks, restart policy, metrics, and persistent volume. The template captures the exact port, public URL, setup-token, cache, persistence, and recovery contract that passed a clean Railway proof.

Backups require a stopped archive of all /data; database-only or live-copy recovery is not promised. AI behavior and third-party provider privacy remain operator/upstream concerns. This independent template supports only its documented Railway packaging contract.