refactor: Remove all LiveKit-specific references

Make the template voice-provider agnostic:
- Renamed 'Voice Agent powered by LiveKit' to 'Conversational Agent'
- Removed LiveKit credential examples from manifest
- Removed LiveKit env var examples from README
- Updated docstrings to be provider-neutral

The agent exposes a standard ACP endpoint that any voice
provider (LiveKit, Twilio, etc.) can connect to.

Author: ben-kluger-scale

src/agentex/lib/cli/templates/voice/README.md.j2

@@ -1,6 +1,6 @@
 # {{ agent_name }} - Voice Agent Template
 
-This is a starter template for building **voice agents** with the AgentEx framework. It provides a pre-configured implementation using the `VoiceAgentBase` class with LiveKit voice integration, conversation state management, guardrails, and streaming support.
+This is a starter template for building **conversational agents** with the AgentEx framework. It provides a pre-configured implementation using the `VoiceAgentBase` class with conversation state management, guardrails, and streaming support.
 
 ## What's Included
 
@@ -10,8 +10,7 @@ This template provides:
 - ✅ **Interruption Handling** - Voice-specific logic for handling user interruptions
 - ✅ **Streaming Support** - Real-time streaming with concurrent guardrail checks
 - ✅ **Guardrail System** - Extensible policy enforcement framework
-- ✅ **LiveKit Ready** - Pre-configured for voice infrastructure
-- ✅ **Gemini Integration** - Uses Gemini 2.5 Flash by default
+- ✅ **Multi-LLM Support** - OpenAI, Azure, Vertex AI, and more
 
 ## What You'll Learn
 
@@ -83,12 +82,6 @@ LLM_MODEL=google/gemini-2.0-flash  # optional
 # Returns canned responses for integration testing
 MOCK_MODE=true
 
-# ============================================================================
-# OPTIONAL: LiveKit (for voice infrastructure)
-# ============================================================================
-# LIVEKIT_URL=wss://your-livekit-instance.com
-# LIVEKIT_API_KEY=your_livekit_api_key
-# LIVEKIT_API_SECRET=your_livekit_api_secret
 ```
 
 **Note**: The manifest is pre-configured with `AGENTEX_BASE_URL: ""` for local testing.

src/agentex/lib/cli/templates/voice/manifest.yaml.j2

@@ -1,6 +1,6 @@
-# Agent Manifest Configuration for Voice Agent
-# ------------------------------------------
-# This file defines how your voice agent should be built and deployed.
+# Agent Manifest Configuration for Conversational Agent
+# -----------------------------------------------------
+# This file defines how your conversational agent should be built and deployed.
 
 # Build Configuration
 # ------------------
@@ -34,7 +34,7 @@ local_development:
 # Agent Configuration
 # -----------------
 agent:
-  acp_type: sync  # Voice agents use sync ACP
+  acp_type: sync  # Conversational agents use sync ACP
 
   # Unique name for your agent
   name: {{ agent_name }}
@@ -43,22 +43,16 @@ agent:
   description: {{ description }}
 
   # Temporal workflow configuration
-  # Voice agents typically don't need Temporal
+  # Conversational agents typically don't need Temporal
   temporal:
     enabled: false
 
-  # Voice-specific credentials
-  # Uncomment and configure if using LiveKit
+  # Credentials for external services
+  # Configure based on your voice provider or other integrations
   credentials: []
-  #   - env_var_name: LIVEKIT_URL
-  #     secret_name: livekit-credentials
-  #     secret_key: url
-  #   - env_var_name: LIVEKIT_API_KEY
-  #     secret_name: livekit-credentials
+  #   - env_var_name: YOUR_API_KEY
+  #     secret_name: your-secret-name
   #     secret_key: api-key
-  #   - env_var_name: LIVEKIT_API_SECRET
-  #     secret_name: livekit-credentials
-  #     secret_key: api-secret
 
   # Environment variables for local development
   # For deployed environments, use Kubernetes secrets (see credentials below)

src/agentex/lib/cli/templates/voice/project/acp.py.j2

@@ -1,10 +1,9 @@
-"""{{ agent_name }} - Voice Agent powered by LiveKit
+"""{{ agent_name }} - Conversational Agent
 
 This agent uses the VoiceAgentBase class which provides:
 - Automatic state management and conversation history
 - Streaming support with interruption handling
 - Guardrail system integration
-- LiveKit voice infrastructure ready
 """
 
 import os

src/agentex/voice/__init__.py

@@ -1,7 +1,7 @@
-"""Voice Agent SDK module for building LiveKit-powered voice agents.
+"""Conversational Agent SDK module for building real-time conversational agents.
 
 This module provides base classes and utilities for creating production-ready
-voice agents with state management, guardrails, and streaming support.
+conversational agents with state management, guardrails, and streaming support.
 
 Example usage:
     from agentex.voice import VoiceAgentBase, AgentState, AgentResponse

src/agentex/voice/agent.py

@@ -1,4 +1,4 @@
-"""Voice Agent base class for LiveKit-powered voice agents.
+"""Conversational Agent base class for real-time conversational agents.
 
 This module provides the VoiceAgentBase class which handles:
 - State management and persistence
@@ -46,7 +46,7 @@
 
 
 class VoiceAgentBase(ABC, Generic[TState, TResponse]):
-    """Base class for voice agents with LiveKit integration.
+    """Base class for conversational agents with streaming support.
     
     This class provides:
     - Automatic state management and persistence