Open WebUI: Add full end-to-end example rig

application/open-webui/README.md

@@ -0,0 +1,64 @@
+# Use CrateDB with Open WebUI
+
+## About
+
+A complete end-to-end rig including CrateDB, CrateDB MCPO, and Open WebUI. 
+
+[Open WebUI] is an extensible, feature-rich, and user-friendly self-hosted AI
+platform designed to operate entirely offline. It supports various LLM runners
+like Ollama and OpenAI-compatible APIs, with built-in inference engine for RAG,
+making it a powerful AI deployment solution.
+
+CrateDB MCPO is a wrapper around CrateDB MCP, publishing the exposed tools via
+OpenAPI, to connect it to Open WebUI.
+
+## What's inside
+
+- `.env`: The dotenv file defines `OPENAI_API_KEY` for `compose.yml` and `OPEN_WEBUI_URL` for `setup.sh`.
+- `compose.yml`: The service composition file defines three services: CrateDB, CrateDB MCPO, and Open WebUI.
+  Use it with `docker compose` or `podman compose`.
+- `setup.Dockerfile`: Container runtime definition for `setup.sh`, including all requirements.
+- `setup.sh`: Program that configures Open WebUI via HTTP API.
+- `tool-servers.json`: JSON snippet for configuring the CrateDB MCPO server through `/api/v1/configs/tool_servers`.
+
+## Operate
+
+Configure the API key for OpenAI as environment variable, or define it within
+the `.env` file.
+```shell
+export OPENAI_API_KEY=<your_openapi_key>
+```
+
+Spin up the software stack.
+```shell
+docker compose up -d
+```
+
+## Usage
+
+You can access the services and their resources on those URLs.
+
+- CrateDB: http://localhost:4200/
+- CrateDB MCPO: http://localhost:5200/docs, http://localhost:5200/openapi.json
+- Open WebUI: http://localhost:6200/, http://localhost:6200/docs
+
+## Configure
+
+To make the ensemble work well, you need to configure a few bits on the Open WebUI
+user interface after ramping up the rig.
+
+- Make sure to enable the "CrateDB" tool on the left side of the query prompt,
+  behind the "More (`+`)" button.
+- Make sure to enable "Function Calling: Native" within the flyout widget about
+  "Chat Controls", to be opened in the top right corner of the screen.
+- If you like, you can dial down to "Temperature: 0.5".
+
+## Example questions
+
+Enjoy conversations with CrateDB and its documentation.
+
+- Text-to-SQL: What is the average value for sensor 1?
+- Knowledgebase: How do I use CrateDB with SQLAlchemy?
+
+
+[Open WebUI]: https://docs.openwebui.com/

application/open-webui/backlog.md

@@ -0,0 +1,6 @@
+# Open WebUI / MCPO backlog
+
+- `compose.yml`: The MCPO OCI currently does not include `curl`.
+  Improve healthcheck after adding it.
+- Add example questions to general instructions à la "If the user asks about questions
+  to use here, enumerate those items: [...]"

application/open-webui/compose.yml

@@ -0,0 +1,124 @@
+# Use CrateDB with Open WebUI
+#
+# https://cratedb.com/docs/
+# https://docs.openwebui.com/getting-started/quick-start
+---
+networks:
+  llm-demo:
+    name: llm-demo
+    driver: bridge
+
+volumes:
+  cratedb:
+  open-webui:
+
+services:
+
+  # -------
+  # CrateDB
+  # -------
+  cratedb:
+    image: docker.io/crate/crate:6.0.0
+    environment:
+      CRATE_HEAP_SIZE: 2g
+    ports:
+      - "4200:4200"
+      - "5432:5432"
+    command: [
+      "crate",
+      "-Cdiscovery.type=single-node",
+      "-Ccluster.routing.allocation.disk.threshold_enabled=false",
+    ]
+    networks:
+      - llm-demo
+    volumes:
+      - cratedb:/data
+    healthcheck:
+      test: [ "CMD", "curl", "--fail", "http://localhost:4200" ]
+      start_period: 3s
+      interval: 10s
+
+  # ------------
+  # CrateDB MCPO
+  # ------------
+  cratedb-mcpo:
+    image: ghcr.io/crate/cratedb-mcpo:0.0.5
+    environment:
+      CRATEDB_CLUSTER_URL: http://crate:crate@cratedb:4200/
+    ports:
+      - "5200:8000"
+    networks:
+      - llm-demo
+    #healthcheck:
+    #  test: [ "CMD", "curl", "--fail", "http://localhost:8000/docs" ]
+    #  start_period: 3s
+    #  interval: 10s
+
+  # ----------
+  # Open WebUI
+  # ----------
+  open-webui:
+    image: ghcr.io/open-webui/open-webui:0.6.18
+    # https://docs.openwebui.com/getting-started/env-configuration
+    # https://docs.openwebui.com/getting-started/api-endpoints/#swagger-documentation-links
+    environment:
+      # From `.env` file.
+      OPENAI_API_KEY:
+      # Defined here.
+      ENABLE_SIGNUP: False
+      ENABLE_LOGIN_FORM: False
+      WEBUI_AUTH: False
+      DEFAULT_MODELS: "gpt-4.1"
+      DEFAULT_USER_ROLE: "admin"
+      ENABLE_CHANNELS: True
+      RESPONSE_WATERMARK: "This text is AI generated"
+      WEBUI_NAME: "CrateDB LLM Cockpit"
+      BYPASS_MODEL_ACCESS_CONTROL: True
+      ENABLE_OLLAMA_API: False
+      ENABLE_OPENAI_API: True
+      ENABLE_DIRECT_CONNECTIONS: True
+      ENV: "dev"
+    ports:
+      - "6200:8080"
+    networks:
+      - llm-demo
+    volumes:
+      - open-webui:/app/backend/data
+    healthcheck:
+      test: [ "CMD", "curl", "--fail", "http://localhost:8080" ]
+      start_period: 3s
+      interval: 10s
+      timeout: 90s
+      retries: 60
+
+  # -----
+  # Setup
+  # -----
+  setup:
+    build:
+      context: .
+      dockerfile: setup.Dockerfile
+    command: bash /app/setup.sh
+    networks:
+      - llm-demo
+    depends_on:
+      open-webui:
+        condition: service_healthy
+
+  # -------
+  # Bundler
+  # -------
+  # Wait for all defined services to be fully available by probing their health
+  # status, even when using `docker compose up --detach`.
+  # https://marcopeg.com/2019/docker-compose-healthcheck/
+  start-dependencies:
+    image: docker.io/dadarek/wait-for-dependencies
+    depends_on:
+      cratedb:
+        condition: service_healthy
+      #cratedb-mcpo:
+      #  condition: service_healthy
+      open-webui:
+        condition: service_healthy
+      setup:
+        condition: service_completed_successfully

application/open-webui/init.sql

@@ -0,0 +1,25 @@
+DROP TABLE IF EXISTS time_series_data;
+
+CREATE TABLE IF NOT EXISTS time_series_data (
+    timestamp TIMESTAMP,
+    value DOUBLE,
+    location STRING,
+    sensor_id INT
+);
+
+INSERT INTO time_series_data (timestamp, value, location, sensor_id)
+VALUES
+    ('2023-09-14T00:00:00', 10.5, 'Sensor A', 1),
+    ('2023-09-14T01:00:00', 15.2, 'Sensor A', 1),
+    ('2023-09-14T02:00:00', 18.9, 'Sensor A', 1),
+    ('2023-09-14T03:00:00', 12.7, 'Sensor B', 2),
+    ('2023-09-14T04:00:00', 17.3, 'Sensor B', 2),
+    ('2023-09-14T05:00:00', 20.1, 'Sensor B', 2),
+    ('2023-09-14T06:00:00', 22.5, 'Sensor A', 1),
+    ('2023-09-14T07:00:00', 18.3, 'Sensor A', 1),
+    ('2023-09-14T08:00:00', 16.8, 'Sensor A', 1),
+    ('2023-09-14T09:00:00', 14.6, 'Sensor B', 2),
+    ('2023-09-14T10:00:00', 13.2, 'Sensor B', 2),
+    ('2023-09-14T11:00:00', 11.7, 'Sensor B', 2);
+
+REFRESH TABLE time_series_data;

application/open-webui/setup.Dockerfile

@@ -0,0 +1,34 @@
+FROM python:3.13-slim-bookworm
+
+# Configure operating system.
+ENV DEBIAN_FRONTEND=noninteractive
+ENV TERM=linux
+
+RUN set -e \
+    && apt-get update \
+    && apt-get --yes install --no-install-recommends --no-install-suggests curl jq \
+    && rm -rf /var/lib/apt/lists/*
+
+# Install and configure `uv`.
+# Guidelines that have been followed.
+# - https://hynek.me/articles/docker-uv/
+
+# Install the `uv` package manager.
+COPY --from=ghcr.io/astral-sh/uv:latest /uv /usr/local/bin/uv
+COPY --from=ghcr.io/astral-sh/uv:latest /uvx /usr/local/bin/uvx
+
+# - Tell uv to byte-compile packages for faster application startups.
+# - Silence uv complaining about not being able to use hard links.
+# - Prevent uv from accidentally downloading isolated Python builds.
+# - Install packages into the system Python environment.
+ENV \
+    UV_COMPILE_BYTECODE=true \
+    UV_LINK_MODE=copy \
+    UV_PYTHON_DOWNLOADS=never \
+    UV_SYSTEM_PYTHON=true
+
+RUN uv pip install crash cratedb-mcp httpie
+
+RUN mkdir /app
+WORKDIR /app
+COPY .env setup.sh *.json *.sql /app/

application/open-webui/setup.sh

@@ -0,0 +1,38 @@
+#!/usr/bin/env bash
+
+# Configure Open WebUI via HTTP API.
+
+# Runtime options.
+set -e
+set -x
+
+# Load variables.
+source .env
+
+# Sign in to receive JWT token.
+token=$( http --ignore-stdin POST ${OPEN_WEBUI_URL}/api/v1/auths/signin email= password= | jq -r .token )
+echo "JWT token: ${token}"
+
+# Inquire health.
+http --ignore-stdin ${OPEN_WEBUI_URL}/health
+
+# Inquire available models.
+#http ${OPEN_WEBUI_URL}/api/models Authorization:"Bearer $token"
+
+# List available tools.
+http --ignore-stdin ${OPEN_WEBUI_URL}/api/v1/tools/ Authorization:"Bearer $token"
+
+# Configure system prompt.
+http --ignore-stdin ${OPEN_WEBUI_URL}/api/v1/users/user/settings/update Authorization:"Bearer $token" ui[system]="$( cratedb-mcp show-prompt )" ui[notificationEnabled]="true"
+
+# Configure CrateDB MCPO server.
+http --ignore-stdin ${OPEN_WEBUI_URL}/api/v1/configs/tool_servers Authorization:"Bearer $token" "@tool-servers.json"
+
+# Configure chat model.
+http --ignore-stdin ${OPEN_WEBUI_URL}/api/v1/configs/models Authorization:"Bearer $token" DEFAULT_MODELS="gpt-4.1" MODEL_ORDER_LIST="[]"
+
+# Configure embedding model.
+http --ignore-stdin ${OPEN_WEBUI_URL}/api/v1/retrieval/embedding/update Authorization:"Bearer $token" embedding_engine="openai" embedding_model="text-embedding-3-small"
+
+# Provision database
+crash --hosts ${CRATEDB_URL} < init.sql

application/open-webui/tool-servers.json

@@ -0,0 +1,18 @@
+{
+  "TOOL_SERVER_CONNECTIONS": [
+    {
+      "url": "http://cratedb-mcpo:8000",
+      "path": "openapi.json",
+      "auth_type": "bearer",
+      "key": "",
+      "config": {
+        "enable": true,
+        "access_control": null
+      },
+      "info": {
+        "name": "CrateDB",
+        "description": "CrateDB Text-to-SQL and documentation inquiry."
+      }
+    }
+  ]
+}