Skip to main content
AllCodex releases regular updates containing bug fixes, new features, and optimized LLM prompts. Upgrading a self-hosted stack takes only a few steps.

Standard Docker Upgrade

Since all of your campaign lore, SQLite files, and vector indices are persisted in /srv/allcodex on your host machine, you can update your Docker containers without losing any data.

1. Pull the Latest Images

Pull the updated tags from GitHub Packages:
docker compose pull

2. Recreate the Containers

Recreate and start the services. Docker will only replace the containers that have new images:
docker compose up -d

3. Clean Up Old Images

Remove old, dangling Docker images to free up disk space:
docker image prune -f

Upgrading Database Schema & Migrations

AllCodex Core (SQLite)

Core applies schema migrations automatically on startup. The Express server checks the SQLite schema, runs any pending sql scripts, and updates the database without manual commands.

AllKnower (PostgreSQL)

If you are running AllKnower via Docker, schema migrations (prisma migrate deploy) are executed automatically by the container entrypoint before launching the web server. If you are running a Local Development setup (Method 2), you must apply Prisma migrations manually after checking out the latest code:
cd allknower
bun db:migrate

Re-indexing Vector Embeddings (RAG)

If a new AllCodex release changes the embedding models or the RAG schema (e.g. updating Qwen embeddings), your LanceDB vector database may need to be re-indexed:
  1. Open the Portal Settings tab.
  2. Under the AI & Search Status card, look for the Reindex RAG section.
  3. Click Reindex All Lore to queue a background job. AllKnower will read all notes tagged with #lore, regenerate embeddings, and rebuild your vector databases.
  4. Alternatively, click Wipe DB Lore & RAG to completely wipe LanceDB and Prisma metadata backends, giving you a fresh start.