Confirmed live: install_frigate()'s fresh-install path overwrote a
working, hand-crafted docker-compose.yml (Frigate + mosquitto +
frigate-notify) with zero backup, because that file's shape didn't match
what frigate.sh's own "existing install" detection knew how to recognize.
Every service's own detection is a judgment call about what counts as
"already installed" and can miss a real setup built outside this repo's
conventions.
lib/common.sh gains backup_if_exists(FILE) — copies FILE to
FILE.bak.<timestamp> if it exists, no-ops otherwise (including DRY_RUN).
Applied before every service's own `cat > docker-compose.yml`/`cat > .env`
write across all 60 services that do one (115 call sites), plus a matching
standalone-mode stub added to every service's own bootstrap block, same
convention already used for port_in_use/find_free_port. This doesn't
replace a service's own update/fresh-reinstall detection — it's the safety
net underneath it, so a wrong detection costs a .bak file to restore from
instead of the original silently disappearing.
Also fixes the actual gap that surfaced this: services/frigate.sh's
Authelia offer only checked for Authelia installed locally on Frigate's
own box, which is never true for a dedicated NVR box with no local Caddy
either (the common shape — Caddy lives elsewhere, snippet-generation mode
already handles that). Now offers Authelia protection unconditionally and,
when Authelia isn't local, asks whether it lives on the same machine as
Caddy (still "import authelia", since that's local to wherever Caddy ends
up) or on a genuinely separate third machine (the explicit
header-pinned forward_auth form, per CLAUDE.md's "forward_auth to a remote
Authelia" note, needed because a bare authelia:9091 shortcut only works
one hop).
Vendor the functional source of github.com/outis1one/local-ai into
./ai-stack (22 files) and add services/ai-stack.sh, which copies the
source to ~/docker/ai-stack and hands off to the app's VRAM-aware
installer (local-ai-setup.sh). The stack bundles Ollama, Open WebUI,
RAG + MCP servers, ChromaDB, SearXNG, Kiwix, Gitea, InvokeAI, ComfyUI
and Portainer.
Cloud LLM providers (Groq/DeepInfra/OpenAI/OpenRouter) are optionally
wired into Open WebUI via the plural OPENAI_API_BASE_URLS list, with the
local RAG connection kept as the first entry so RAG keeps working. Open
WebUI ships built-in auth, so Caddy is configured without Authelia.
Excludes the upstream's two bundled copies of this very project
(ubuntu-post-install.sh, ubuntu-post-install-main.zip) — stale and
circular. Coexists with the existing ai-gpu service.
Also fix the install-function names for ai-gpu and ai-stack: the
dispatcher calls install_<raw-name>, so the function must be
install_ai-gpu / install_ai-stack (hyphen), matching the working
mail-archiver / wg-easy services. ai-gpu was previously uninstallable.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Nb2vJ8W7bHKx1JXVvpCraH
Generalize the optional Groq toggle into a provider picker so the GPU AI
stack can wire Groq, DeepInfra, OpenAI, and/or OpenRouter into Open WebUI
at once. All are OpenAI-compatible, so they share Open WebUI's
OPENAI_API_BASE_URLS / OPENAI_API_KEYS (semicolon-separated, matched by
position). Keys are written to llm/.env; only ${VAR} references go into
docker-compose.yml.
Updates the install prompt, .env template, compose patch, README, the
dry-run note, and the final summary accordingly.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Nb2vJ8W7bHKx1JXVvpCraH
Wires Groq (api.groq.com) into Open WebUI as an OpenAI-compatible
connection so its hosted models show up alongside local Ollama models.
Prompts for an API key during install, stores it in llm/.env, and
patches the cloned docker-compose.yml's open-webui environment block
(idempotently) to reference it.
iopaint:
- Add model selection menu (10 choices) with CPU/GPU/SD tiers and
size/use-case descriptions shown at install time
- Fix volume mount: ./models:/root/.cache (was only /root/.cache/iopaint)
— now persists both torch hub cache (LaMa) and HuggingFace cache (SD/PowerPaint)
- Refactor compose to use ${MODEL} and ${DEVICE} env vars so switching
models only requires editing .env + restart, no compose file edit needed
- Add PowerPaint-V2-filling and SD 1.5 inpainting as explicit menu choices
for text-guided object replacement
- Update header and README to document all three use cases (erase, fill, replace)
and note that IOPaint is local-only (cannot use a remote GPU)
ai-gpu:
- Add Ollama model selection menu (8 models, multi-select with sizes/descriptions)
defaulting to llama3.2:3b + nomic-embed-text
- Auto-pull selected Ollama models immediately after LLM stack starts
- Add InvokeAI starter model selection (SD 1.5 / SDXL Turbo / SDXL Base / skip)
- Queue InvokeAI model download via REST API (POST /api/v2/models/install)
with fallback instructions if the API is unavailable
- Add HuggingFace token prompt; stored as HUGGING_FACE_HUB_TOKEN in image-gen .env
- Wire SearXNG into Open WebUI via ENABLE_RAG_WEB_SEARCH + SEARXNG_QUERY_URL in llm .env
- Default start choice is now 2 (portal + LLM + Ollama pull) so the stack
is ready to use immediately after install
https://claude.ai/code/session_01JEu7LgCWXKhXo18MeYFRZp
iopaint: AI image inpainting (object removal, fill, restore) via IOPaint +
LaMa model. Runs CPU by default; GPU option writes nvidia deploy block.
No built-in auth — Authelia SSO prompt included. Port 8100.
ai-gpu: GPU AI stack from outis1one/ai-6gb-gpu. Clones repo and sets up
three stacks under ~/docker/ai-gpu/: InvokeAI image gen (port 9090),
Ollama + Open WebUI + SearXNG LLM stack (ports 11434/3000), and Flask
portal (port 8080) that hot-swaps the GPU between stacks. Patches
hardcoded home paths in portal docker-compose.yml to use ACTUAL_HOME.
Prompts for TZ (replaces hardcoded America/New_York). Caddy for both
portal (localai) and InvokeAI (images).
https://claude.ai/code/session_01JEu7LgCWXKhXo18MeYFRZp
Claude