Release LiveContext CE v0.1.11 · livecontext-ai/livecontext-ce

What is new

Fresh installs now start with the current AI model catalog. Each release bakes the cloud's signed model bundle into catalog-seeds/model-bundle.json, and the app verifies and applies it at first boot, no cloud link needed. Previous releases (v0.1.9/v0.1.10) shipped without that file due to an export bug, so new installs only saw an outdated built-in model list. Existing installs pick it up with git pull + docker compose up -d (the bundle is applied at the next start; your own model edits are preserved). Continuous updates between releases remain a benefit of linking the install to LiveContext Cloud.
Removed the billing 404 noise on every page. Community Edition no longer calls the cloud-only PAYG billing endpoint, which fired a failed request and an "Error fetching PAYG tiers" console error on every page navigation right after first login.
Cloud AI relay reliability (linked installs). Relayed model calls now always execute on the real provider API; a platform-side execution-link configuration can no longer break relayed models with an opaque 400 error.

The release pipeline now fails loudly if a release would ship without its model refresh (pre-tag publish test, post-publish verification, and a boot-time apply assertion in the release smoke test).

git pull
docker compose pull
docker compose up -d

The compose file pins this release's image tag, so run git pull first. Your data lives in Docker volumes and is kept across updates.

git clone https://github.com/livecontext-ai/livecontext-ce
cd livecontext-ce
docker compose up -d

ghcr.io/livecontext-ai/livecontext-ce:v0.1.11
ghcr.io/livecontext-ai/livecontext-ce-frontend:v0.1.11
ghcr.io/livecontext-ai/livecontext-ce-bridge:v0.1.11
ghcr.io/livecontext-ai/livecontext-ce-screenshot-renderer:v0.1.11 (opt-in renderer profile)