Deploy and Host Grimoire Bookmark Workspace

Local-first bookmarks and search in a private persistent workspace.

Validation scope: source review and static checks. This template meets the selection criterion through a public marketplace gap: no matching listing was found on September 19, 2026 using Grimoire, grimoire bookmarks, littleimp, little imp, goniszewski. Searches are bounded; private, unindexed and differently named listings may exist. The selection does not claim any competitor is broken.

About Hosting

Service	Network	Persistent mount
core	Private	/data
grimoire	Public HTTPS	None

The public gateway requires username admin and ACCESS_PASSWORD from the grimoire service. API clients can send X-Template-Key: YOUR_ACCESS_PASSWORD. Keep core and databases private. The gateway strips Basic Authorization, preserves Bearer authorization and WebSocket upgrades, and limits requests to 32 MiB. Verify native client compatibility before relying on it.

Why Deploy Grimoire Bookmark Workspace on Railway

This template wires a protected HTTPS entry point to private application services, uses generated owner credentials, and declares the persistent mounts shown above. Repository adapters and pinned image references keep the deployment configuration reviewable. Application setup and the acceptance checks below remain operator responsibilities.

Common Use Cases

Collect research bookmarks; search saved articles; optionally enrich bookmarks with your own AI provider.

First use

Open the workspace behind the gateway. Optional AI enrichment requires your own model credentials in the application settings. The current v1.2 rewrite is built from a checksum-verified source archive.

Recommended acceptance checks

Save a URL, import a small bookmark file, extract readable content, search it and verify data after restart. Test semantic search separately if a provider is configured.

Test a volume-preserving redeploy as well as a restart. For stateful apps, take an application-consistent backup, including database/WAL and uploaded files, restore it into a separate disposable instance and verify account access and records. For configuration-as-code services, verify a clean rebuild from the saved source. Keep the original data until restoration passes.

Scope and limitations

Not an upgrade path for legacy v0.5 installations. AI providers cost extra. Build dependencies and the application workflow have not been executed in Railway.

This template has source/configuration review and static checks only. No container build, Railway startup, browser/API workflow, volume recovery or cost measurement was performed. Saved editor fidelity is not deployment proof; /healthz proves only gateway readiness. Set operator-owned provider credentials only where needed and inspect the selected upstream license before commercial use.

Dependencies for Grimoire Bookmark Workspace

Deployment Dependencies

A Railway account with capacity for the 2 services listed above, access to the selected image registries and GitHub source branch, and persistent volumes where shown are required. Keep volume-backed services at one replica. Provider accounts, SMTP and other optional integrations are supplied by the operator as described in First use and Scope and limitations.

Upstream sources

Upstream project
Selected source reference
Upstream source receipt: ca974f3561399302b56e2b2b7f52036766dc4117.
Images are digest-pinned in images.round8.lock.json. Rolling image tags are frozen at the captured digest, not asserted to match the source commit unless independently established.
Adapters and the gateway use orenaksakal/railway-templates, branch codex/unique-template-drafts, with repository-root Docker build context. Keep the branch available.
Configuration reference
Configuration reference
Configuration reference