Split GPU bits into compose.gpu.yml override (#10)#80
Merged
Conversation
The remaining piece of #15 (the in-repo half — bridge code hardening already shipped in aa2d8ba) was that fresh deployments of zeroclaw-bridge had no obvious place to put OPENROUTER_API_KEY / VISION_API_KEY / VLM_API_KEY, so they'd silently start with an empty key and the bridge's ERROR fallback would fire on every photo intent. - `zeroclaw-bridge.service.template`: add `EnvironmentFile=-/root/zeroclaw-bridge/.env`. The leading `-` keeps systemd from failing startup if the file is absent — the bridge will surface the missing-key ERROR loudly on the first photo call instead. - `scripts/install-bridge.sh`: same `EnvironmentFile=-${BRIDGE_DIR}/.env` in the inline unit, plus creation of a mode-0600 stub `.env` with `OPENROUTER_API_KEY=` (commented `VISION_API_KEY` / `VLM_API_KEY` for the split-key case). Existing `.env` is preserved. This doesn't retroactively fix the production RPi mentioned in the issue body — that one needs `systemctl edit zeroclaw-bridge` or a manual `.env` drop — but it ensures every new install lands with the env file present and the location obvious. Becomes moot when #36 retires the RPi.
BrettKinny
force-pushed
the
claude/address-open-issues-xvDUe
branch
from
85d08ea to
8d258e9
Compare
There was a problem hiding this comment.
Pull request overview
This PR makes the Docker stack CPU-portable by moving NVIDIA-specific Compose configuration into an optional GPU override, while also introducing an idempotent setup wizard and bridge .env loading for API keys.
Changes:
- Adds
compose.gpu.ymland updates Compose templates/docs to make GPU support opt-in. - Replaces inline Makefile setup logic with
scripts/setup-wizard.sh. - Adds bridge
.envloading/stub creation and changelog/gitignore updates for generated config files.
Reviewed changes
Copilot reviewed 3 out of 3 changed files in this pull request and generated 6 comments.
Show a summary per file
| File | Description |
|---|---|
zeroclaw-bridge.service.template |
Loads optional bridge runtime environment file. |
scripts/setup-wizard.sh |
New setup/regeneration wizard with GPU detection and template rendering. |
scripts/install-bridge.sh |
Adds bridge .env systemd loading and stub creation. |
Makefile |
Delegates setup to the wizard and adds GPU/ASR doctor check. |
docker-compose.yml.template |
Removes hard-coded GPU settings and uses timezone placeholder. |
compose.gpu.yml |
Adds optional NVIDIA runtime/device override. |
CHANGELOG.md |
Documents setup, GPU split, and bridge env changes. |
.gitignore |
Ignores generated config/runtime artifacts. |
.env.example |
Documents optional COMPOSE_FILE GPU override. |
.config.yaml.template |
Adds timezone/ASR/persona placeholders. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
make setupwas failing on hosts without the NVIDIA Container Toolkitbecause
docker-compose.ymlhard-codedruntime: nvidiaplus a GPUreservation block. Daemon rejected the file with "unknown or invalid
runtime name: nvidia" before any container could start.
Moves the GPU bits (
runtime: nvidia,deploy.resources.reservations,NVIDIA_*env vars) into a new optionalcompose.gpu.ymloverride sothe base compose is CPU-portable. The wizard auto-detects the nvidia
runtime via
docker infoand either:COMPOSE_FILE=docker-compose.yml:compose.gpu.ymlinto
.envso alldocker composeinvocations pick up the override.selected_module.ASRfromWhisperLocaltoFunASR(the CPU-friendly alternative already configured in.config.yaml) and removes any stale override line.Adds a
make doctorcheck that flags WhisperLocal+device: cudaconfigured against a host with no nvidia runtime, and documents the
override in
.env.example.