Documentation
Reference for running, configuring, and integrating Trove KB at v0.1.0. The source of truth is the docs/ folder in the app repository; this is a snapshot per release.
- Getting started
Clone, configure three values, docker compose up, create the first administrator.
- Environment variables
Every variable in .env.example, what it does, and its default.
- REST API and webhooks
/api/v1 with bearer keys and an OpenAPI spec, company scoping, PSA lookups, deep links, signed webhooks.
- MCP for AI assistants
POST /api/mcp on the same API keys. Read documentation and the knowledge base; never a secret, never a reveal tool.
- Knowledge base
Collections from outside sources, imports and connectors, runbooks with stable step ids, the public site, grants, and the /kb API.
- Domain checks
DNS, TLS, RDAP, mail posture, and website branding per domain record, on request and on a schedule in three layers, with webhooks for what changed and what is expiring.
- Vault integration
Bitwarden, Vaultwarden, 1Password Connect, and HashiCorp KV. Trove KB stores references and brokers access; it never stores a secret.
- Cloudflare
The container behind a Cloudflare Tunnel, the public knowledge base on its own hostname, or the whole app as a Worker.
- Backup and restore
pg_dump the database, archive the uploads volume, keep .env. A Trove KB backup never contains a secret.
- Accounts, roles, and access
Roles, per-company access, password policy, second factors, single sign-on, and the audit log.
- Starter doc types
What SEED_ON_START loads, field by field, with the shared option lists.