feat: add custom base URL support for local AI services

Add comprehensive support for connecting to local and custom AI endpoints like LM Studio, vLLM, Ollama, and other OpenAI-compatible services.

Changes:

1. Configuration options (config.yaml):
   - Added ai-base-url for primary provider
   - Added ai0-base-url through ai9-base-url for additional slots
   - Updated all provider descriptions with full list of 15 supported providers
   - Added GitHub Copilot provider support
   - Enhanced api-key documentation with provider-specific details

2. Config generation (hooks/common.sh):
   - Implemented models.providers section generation in openclaw.json
   - Dynamically collects custom baseUrls from primary and ai0-ai9 slots
   - Generates proper JSON structure only when baseUrls are configured
   - Handles multiple providers with different custom endpoints
   - Changed auth profile type from api_key to token for OpenClaw 2026.x
   - Removed legacy environment variable API key storage (now uses auth-profiles.json)

3. Documentation (README.md):
   - Added examples for LM Studio, vLLM, Text Generation WebUI
   - Updated Multi-AI Model Support section with baseUrl parameters
   - Added Using Custom Base URLs with Multiple Models section
   - Documented mixing cloud and local AI models simultaneously

Technical details:
- Research confirmed baseUrl belongs in openclaw.json at models.providers.<provider>.baseUrl
- Auth-profiles.json only handles authentication (type, provider, key/token)
- Supports up to 11 AI models with custom endpoints (1 primary + 10 slots)
- Backward compatible - existing configs without baseUrls work unchanged

Examples:
- LM Studio: ai-provider=openai ai-base-url=http://localhost:1234/v1
- vLLM: ai-provider=openai ai-base-url=http://localhost:8000/v1
- Mix cloud + local: primary=anthropic, slot0=LM Studio, slot1=vLLM

Author: fourdollars

README.md

@@ -189,6 +189,34 @@ juju config openclaw \
   ai-model="llama3"
 ```
 
+**Local Models (LM Studio)**
+
+```bash
+juju config openclaw \
+  ai-provider="openai" \
+  ai-api-key="lm-studio" \
+  ai-model="local-model" \
+  ai-base-url="http://localhost:1234/v1"
+```
+
+**Other OpenAI-compatible Services**
+
+```bash
+# vLLM
+juju config openclaw \
+  ai-provider="openai" \
+  ai-api-key="dummy-key" \
+  ai-model="meta-llama/Llama-3.3-70B-Instruct" \
+  ai-base-url="http://localhost:8000/v1"
+
+# Text Generation WebUI
+juju config openclaw \
+  ai-provider="openai" \
+  ai-api-key="dummy-key" \
+  ai-model="local-model" \
+  ai-base-url="http://localhost:5000/v1"
+```
+
 ### Multi-AI Model Support
 
 OpenClaw Charm supports configuring up to 11 AI models simultaneously (1 primary + 10 additional slots). This enables model switching, fallback, and specialized model usage for different tasks.
@@ -221,19 +249,50 @@ juju config openclaw \
 
 **All AI Slots:**
 
-- Primary: `ai-provider`, `ai-model`, `ai-api-key`
-- Slot 0: `ai0-provider`, `ai0-model`, `ai0-api-key`
-- Slot 1: `ai1-provider`, `ai1-model`, `ai1-api-key`
-- Slot 2: `ai2-provider`, `ai2-model`, `ai2-api-key`
-- Slot 3: `ai3-provider`, `ai3-model`, `ai3-api-key`
-- Slot 4: `ai4-provider`, `ai4-model`, `ai4-api-key`
-- Slot 5: `ai5-provider`, `ai5-model`, `ai5-api-key`
-- Slot 6: `ai6-provider`, `ai6-model`, `ai6-api-key`
-- Slot 7: `ai7-provider`, `ai7-model`, `ai7-api-key`
-- Slot 8: `ai8-provider`, `ai8-model`, `ai8-api-key`
-- Slot 9: `ai9-provider`, `ai9-model`, `ai9-api-key`
+- Primary: `ai-provider`, `ai-model`, `ai-api-key`, `ai-base-url` (optional)
+- Slot 0: `ai0-provider`, `ai0-model`, `ai0-api-key`, `ai0-base-url` (optional)
+- Slot 1: `ai1-provider`, `ai1-model`, `ai1-api-key`, `ai1-base-url` (optional)
+- Slot 2: `ai2-provider`, `ai2-model`, `ai2-api-key`, `ai2-base-url` (optional)
+- Slot 3: `ai3-provider`, `ai3-model`, `ai3-api-key`, `ai3-base-url` (optional)
+- Slot 4: `ai4-provider`, `ai4-model`, `ai4-api-key`, `ai4-base-url` (optional)
+- Slot 5: `ai5-provider`, `ai5-model`, `ai5-api-key`, `ai5-base-url` (optional)
+- Slot 6: `ai6-provider`, `ai6-model`, `ai6-api-key`, `ai6-base-url` (optional)
+- Slot 7: `ai7-provider`, `ai7-model`, `ai7-api-key`, `ai7-base-url` (optional)
+- Slot 8: `ai8-provider`, `ai8-model`, `ai8-api-key`, `ai8-base-url` (optional)
+- Slot 9: `ai9-provider`, `ai9-model`, `ai9-api-key`, `ai9-base-url` (optional)
+
+**Note:** Each slot requires at minimum the three core parameters (provider, model, api-key). The `base-url` parameter is optional and only needed for custom API endpoints. Partially configured slots will trigger validation errors.
+
+**Using Custom Base URLs with Multiple Models:**
+
+```bash
+# Mix cloud and local AI models
+juju config openclaw \
+  ai-provider="anthropic" \
+  ai-api-key="sk-ant-xxx" \
+  ai-model="claude-opus-4-5"
+
+# Add LM Studio instance in slot 0
+juju config openclaw \
+  ai0-provider="openai" \
+  ai0-api-key="lm-studio" \
+  ai0-model="local-llama3" \
+  ai0-base-url="http://localhost:1234/v1"
+
+# Add vLLM instance in slot 1
+juju config openclaw \
+  ai1-provider="openai" \
+  ai1-api-key="vllm-key" \
+  ai1-model="meta-llama/Llama-3.3-70B-Instruct" \
+  ai1-base-url="http://192.168.1.100:8000/v1"
+```
+
+This configuration creates:
+- Primary model using Anthropic's cloud API
+- Slot 0 using LM Studio running locally
+- Slot 1 using a vLLM server on your network
 
-**Note:** Each slot requires all three parameters (provider, model, api-key) to be configured. Partially configured slots will trigger validation errors.
+OpenClaw will generate the appropriate `models.providers` configuration in `openclaw.json` with custom baseUrls for each provider that has one specified.
 
 ### Messaging Channels
 
@@ -320,9 +379,11 @@ juju config openclaw enable-browser-tool=true
 | `ai-provider` | string | - | AI provider: anthropic, openai, google, bedrock, ollama |
 | `ai-model` | string | - | AI model name |
 | `ai-api-key` | string | - | API key for selected provider |
+| `ai-base-url` | string | - | Custom API base URL (for LM Studio, vLLM, etc.) |
 | `ai0-provider` ... `ai9-provider` | string | - | Additional AI provider slots (0-9) |
 | `ai0-model` ... `ai9-model` | string | - | Additional AI model name slots (0-9) |
 | `ai0-api-key` ... `ai9-api-key` | string | - | Additional AI API key slots (0-9) |
+| `ai0-base-url` ... `ai9-base-url` | string | - | Custom API base URL for additional slots (0-9) |
 | `telegram-bot-token` | string | - | Telegram bot token from @BotFather |
 | `discord-bot-token` | string | - | Discord bot token |
 | `slack-bot-token` | string | - | Slack bot token (xoxb-...) |

config.yaml

@@ -23,7 +23,22 @@ options:
     type: string
     default: ""
     description: |
-      Primary AI model provider: 'anthropic', 'openai', 'google', 'bedrock', or 'ollama'
+      Primary AI model provider. Supported providers:
+      - anthropic: Anthropic Claude models
+      - openai: OpenAI GPT models  
+      - openai-codex: OpenAI Codex (requires OAuth)
+      - google: Google Gemini models
+      - opencode: OpenCode Zen models
+      - github-copilot: GitHub Copilot/Models API
+      - openrouter: OpenRouter aggregator
+      - xai: xAI models
+      - groq: Groq models
+      - cerebras: Cerebras models
+      - mistral: Mistral AI models
+      - zai: Z.AI (GLM) models
+      - vercel-ai-gateway: Vercel AI Gateway
+      - ollama: Local Ollama models (no API key needed)
+      - bedrock: AWS Bedrock (uses AWS credentials)
   
   ai-model:
     type: string
@@ -34,16 +49,41 @@ options:
     type: string
     default: ""
     description: |
-      API key for the selected AI provider (anthropic, openai, or google).
-      For Anthropic: Anthropic API key
-      For OpenAI: OpenAI API key
-      For Google: Google Gemini API key
+      API key for the selected AI provider. Required for most providers:
+      - anthropic: Anthropic API key
+      - openai: OpenAI API key
+      - google: Google Gemini API key
+      - opencode: OpenCode API key
+      - github-copilot: GitHub personal access token
+      - openrouter: OpenRouter API key
+      - xai: xAI API key
+      - groq: Groq API key
+      - cerebras: Cerebras API key
+      - mistral: Mistral API key
+      - zai: Z.AI API key
+      - vercel-ai-gateway: Vercel AI Gateway API key
+      Not required for: ollama (local), bedrock (AWS credentials), openai-codex (OAuth)
+      
+      Note: API keys are stored securely in auth-profiles.json, not in environment variables.
+  
+  ai-base-url:
+    type: string
+    default: ""
+    description: |
+      Custom API base URL for OpenAI-compatible providers. 
+      Use this to connect to local AI services like LM Studio, vLLM, FastChat, etc.
+      Examples:
+      - LM Studio: http://localhost:1234/v1
+      - vLLM: http://localhost:8000/v1
+      - Text Generation WebUI: http://localhost:5000/v1
+      Leave empty to use default provider endpoints.
   
   # Additional AI models (ai0 through ai9) for multi-model support
   ai0-provider:
     type: string
     default: ""
-    description: AI provider for model slot 0 (anthropic, openai, google, bedrock, ollama)
+    description: |
+      AI provider for model slot 0. See ai-provider for supported providers.
   
   ai0-model:
     type: string
@@ -55,10 +95,15 @@ options:
     default: ""
     description: API key for AI model slot 0
 
+  ai0-base-url:
+    type: string
+    default: ""
+    description: Custom API base URL for AI model slot 0 (for OpenAI-compatible local services)
+  
   ai1-provider:
     type: string
     default: ""
-    description: AI provider for model slot 1 (anthropic, openai, google, bedrock, ollama)
+    description: AI provider for model slot 1. See ai-provider for supported providers.
 
   ai1-model:
     type: string
@@ -70,10 +115,15 @@ options:
     default: ""
     description: API key for AI model slot 1
 
+  ai1-base-url:
+    type: string
+    default: ""
+    description: Custom API base URL for AI model slot 1 (for OpenAI-compatible local services)
+  
   ai2-provider:
     type: string
     default: ""
-    description: AI provider for model slot 2 (anthropic, openai, google, bedrock, ollama)
+    description: AI provider for model slot 2. See ai-provider for supported providers.
 
   ai2-model:
     type: string
@@ -85,10 +135,15 @@ options:
     default: ""
     description: API key for AI model slot 2
 
+  ai2-base-url:
+    type: string
+    default: ""
+    description: Custom API base URL for AI model slot 2 (for OpenAI-compatible local services)
+  
   ai3-provider:
     type: string
     default: ""
-    description: AI provider for model slot 3 (anthropic, openai, google, bedrock, ollama)
+    description: AI provider for model slot 3. See ai-provider for supported providers.
 
   ai3-model:
     type: string
@@ -100,10 +155,15 @@ options:
     default: ""
     description: API key for AI model slot 3
 
+  ai3-base-url:
+    type: string
+    default: ""
+    description: Custom API base URL for AI model slot 3 (for OpenAI-compatible local services)
+  
   ai4-provider:
     type: string
     default: ""
-    description: AI provider for model slot 4 (anthropic, openai, google, bedrock, ollama)
+    description: AI provider for model slot 4. See ai-provider for supported providers.
 
   ai4-model:
     type: string
@@ -115,10 +175,15 @@ options:
     default: ""
     description: API key for AI model slot 4
 
+  ai4-base-url:
+    type: string
+    default: ""
+    description: Custom API base URL for AI model slot 4 (for OpenAI-compatible local services)
+  
   ai5-provider:
     type: string
     default: ""
-    description: AI provider for model slot 5 (anthropic, openai, google, bedrock, ollama)
+    description: AI provider for model slot 5. See ai-provider for supported providers.
 
   ai5-model:
     type: string
@@ -130,10 +195,15 @@ options:
     default: ""
     description: API key for AI model slot 5
 
+  ai5-base-url:
+    type: string
+    default: ""
+    description: Custom API base URL for AI model slot 5 (for OpenAI-compatible local services)
+  
   ai6-provider:
     type: string
     default: ""
-    description: AI provider for model slot 6 (anthropic, openai, google, bedrock, ollama)
+    description: AI provider for model slot 6. See ai-provider for supported providers.
 
   ai6-model:
     type: string
@@ -145,10 +215,15 @@ options:
     default: ""
     description: API key for AI model slot 6
 
+  ai6-base-url:
+    type: string
+    default: ""
+    description: Custom API base URL for AI model slot 6 (for OpenAI-compatible local services)
+  
   ai7-provider:
     type: string
     default: ""
-    description: AI provider for model slot 7 (anthropic, openai, google, bedrock, ollama)
+    description: AI provider for model slot 7. See ai-provider for supported providers.
 
   ai7-model:
     type: string
@@ -160,10 +235,15 @@ options:
     default: ""
     description: API key for AI model slot 7
 
+  ai7-base-url:
+    type: string
+    default: ""
+    description: Custom API base URL for AI model slot 7 (for OpenAI-compatible local services)
+  
   ai8-provider:
     type: string
     default: ""
-    description: AI provider for model slot 8 (anthropic, openai, google, bedrock, ollama)
+    description: AI provider for model slot 8. See ai-provider for supported providers.
 
   ai8-model:
     type: string
@@ -175,10 +255,15 @@ options:
     default: ""
     description: API key for AI model slot 8
 
+  ai8-base-url:
+    type: string
+    default: ""
+    description: Custom API base URL for AI model slot 8 (for OpenAI-compatible local services)
+  
   ai9-provider:
     type: string
     default: ""
-    description: AI provider for model slot 9 (anthropic, openai, google, bedrock, ollama)
+    description: AI provider for model slot 9. See ai-provider for supported providers.
 
   ai9-model:
     type: string
@@ -190,6 +275,11 @@ options:
     default: ""
     description: API key for AI model slot 9
 
+  ai9-base-url:
+    type: string
+    default: ""
+    description: Custom API base URL for AI model slot 9 (for OpenAI-compatible local services)
+  
   telegram-bot-token:
     type: string
     default: ""