@edgible-team/cli 1.0.1 → 1.2.0

package/LICENSE

@@ -134,3 +134,19 @@ software under these terms.
 **Use** means anything you do with the software requiring one
 of your licenses.
 
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+

package/README.md

@@ -301,8 +301,36 @@ edgible ai stop
 
 | Command | Description |
 |---------|-------------|
-| `edgible ai setup` | Setup Ollama AI service |
+| `edgible ai setup` | Complete AI setup wizard (local + optional platform integration) |
 | `edgible ai stop` | Stop Ollama service |
+| `edgible ai serve` | Start Open WebUI connected to local Ollama |
+| `edgible ai teardown` | Stop services and optionally remove platform applications |
+| `edgible ai status` | Check status of AI services |
+| `edgible ai test [--model]` | Test model connectivity and response |
+
+#### AI Setup Options
+
+| Option | Description |
+|--------|-------------|
+| `--model <model>` | Specify model to use (skips interactive selection) |
+| `--auto-install` | Automatically install Ollama without prompting |
+| `--local-only` | Skip platform integration (local setup only) |
+| `--expose-ollama` | Expose Ollama API publicly via HTTPS |
+| `--setup-webui` | Setup Open WebUI interface |
+| `--webui-deployment <type>` | WebUI deployment: `local` or `remote` |
+| `--device-id <id>` | Serving device ID for both apps |
+| `--ollama-device-id <id>` | Serving device ID for Ollama |
+| `--webui-device-id <id>` | Serving device ID for Open WebUI |
+| `--gateway-ids <ids>` | Comma-separated gateway device IDs |
+| `--no-interactive` | Run in non-interactive mode |
+
+#### AI Teardown Options
+
+| Option | Description |
+|--------|-------------|
+| `--stop-ollama` | Also stop the Ollama service |
+| `--remove-volumes` | Remove Docker data volumes |
+| `--remove-apps` | Remove platform applications (ollama-api, open-webui) |
 
 ## Examples
 
@@ -349,6 +377,91 @@ edgible agent status
 edgible agent logs --follow
 ```
 
+### Set Up AI with Ollama and Open WebUI
+
+The `edgible ai setup` command provides an interactive wizard for complete AI setup:
+
+#### Option 1: Local Only Setup
+
+```bash
+# Install Ollama, pull model, skip platform integration
+edgible ai setup --local-only
+
+# Manually start Open WebUI later
+edgible ai serve
+```
+
+#### Option 2: Platform-Integrated Setup (Recommended)
+
+```bash
+# Interactive wizard - guides through everything
+edgible ai setup
+
+# The wizard will:
+# 1. Install Ollama locally
+# 2. Detect system capabilities
+# 3. Recommend and pull a model
+# 4. Test model connectivity
+# 5. Ask: Expose Ollama API publicly? (creates HTTPS endpoint)
+# 6. Ask: Setup Open WebUI?
+#    - Local (Docker on this machine)
+#    - Remote (As platform application)
+# 7. Create applications and display URLs
+```
+
+#### Manual Application Creation (Alternative)
+
+```bash
+# 1. Setup Ollama locally first
+edgible ai setup --local-only
+
+# 2. Manually create Ollama API application
+edgible application create existing \
+  --name ollama-api \
+  --port 11434 \
+  --protocol https \
+  --device-id <serving-device> \
+  --gateway-ids <gateway-id>
+
+# 3. Manually create Open WebUI application
+edgible application create existing \
+  --name open-webui \
+  --port 3200 \
+  --protocol https \
+  --device-id <serving-device> \
+  --gateway-ids <gateway-id>
+```
+
+#### Non-Interactive Setup
+
+```bash
+# Complete setup without prompts
+edgible ai setup \
+  --model llama3.2:3b \
+  --expose-ollama \
+  --setup-webui \
+  --webui-deployment remote \
+  --device-id <device-id> \
+  --gateway-ids <gateway-id> \
+  --no-interactive
+```
+
+#### Management Commands
+
+```bash
+# Check status
+edgible ai status
+
+# Test model connectivity
+edgible ai test --model llama3.2:3b
+
+# Teardown everything
+edgible ai teardown --stop-ollama --remove-apps --remove-volumes
+
+# List all applications
+edgible application list
+```
+
 ## Output Options
 
 The CLI supports several output options: