pixelle-mcp-Image-generation

# Copyright (C) 2025 AIDC-AI # This project is licensed under the MIT License (SPDX-License-identifier: MIT). """LLM provider manager.""" from typing import Dict, List, Optional import questionary from rich.console import Console from rich.panel import Panel from pixelle.cli.setup.providers.openai import configure_openai from pixelle.cli.setup.providers.ollama import configure_ollama from pixelle.cli.setup.providers.gemini import configure_gemini from pixelle.cli.setup.providers.deepseek import configure_deepseek from pixelle.cli.setup.providers.claude import configure_claude from pixelle.cli.setup.providers.qwen import configure_qwen console = Console() def configure_specific_llm(provider: str) -> Optional[Dict]: """Configure specific LLM provider""" if provider == "openai": return configure_openai() elif provider == "ollama": return configure_ollama() elif provider == "gemini": return configure_gemini() elif provider == "deepseek": return configure_deepseek() elif provider == "claude": return configure_claude() elif provider == "qwen": return configure_qwen() return None def setup_multiple_llm_providers() -> Optional[List[Dict]]: """Setup multiple LLM providers - Step 2""" console.print(Panel( "🤖 [bold]LLM provider configuration[/bold]\n\n" "Pixelle MCP supports multiple LLM providers, you can configure one or more.\n" "The benefits of configuring multiple providers:\n" "• Can use different models in different scenarios\n" "• Provide backup solutions, improve service availability\n" "• Some models perform better on specific tasks", title="Step 2/4: LLM provider configuration", border_style="green" )) configured_providers = [] while True: # Show available providers available_providers = [ questionary.Choice("🔥 OpenAI (recommended) - GPT-4, GPT-3.5, etc.", "openai"), questionary.Choice("🏠 Ollama (local) - Free local model", "ollama"), questionary.Choice("💎 Google Gemini - Google latest model", "gemini"), questionary.Choice("🚀 DeepSeek - High-performance code model", "deepseek"), questionary.Choice("🤖 Claude - Anthropic's powerful model", "claude"), questionary.Choice("🌟 Qwen - Alibaba Tongyi Qwen", "qwen"), ] # Filter configured providers remaining_providers = [p for p in available_providers if p.value not in [cp["provider"] for cp in configured_providers]] if not remaining_providers: console.print("✅ All available LLM providers are configured, automatically enter next step") break # Show currently configured providers if configured_providers: console.print("\n📋 [bold]Configured providers:[/bold]") for provider in configured_providers: console.print(f" ✅ {provider['provider'].title()}") # Select provider to configure if configured_providers: remaining_providers.append(questionary.Choice("🏁 Complete configuration", "done")) # Always add exit option remaining_providers.append(questionary.Choice("❌ Cancel configuration", "cancel")) provider = questionary.select( "Select LLM provider to configure:" if not configured_providers else "Select LLM provider to continue configuration:", choices=remaining_providers ).ask() if provider is None: # User pressed Ctrl+C (questionary returns None) return None if provider == "cancel": cancel_confirm = questionary.confirm("Are you sure you want to cancel configuration?", default=False, instruction="(y/N)").ask() if cancel_confirm is None: # User pressed Ctrl+C during confirmation return None if cancel_confirm: console.print("❌ Configuration cancelled") return None else: continue # Continue configuration loop if provider == "done": break # Configure specific provider provider_config = configure_specific_llm(provider) if provider_config: configured_providers.append(provider_config) # Show selected models models = provider_config.get('models', '') if models: model_list = [m.strip() for m in models.split(',')] model_display = '、'.join(model_list) console.print(f"✅ [bold green]{provider.title()} configuration successful![/bold green]") console.print(f"📋 You selected {model_display} model\n") else: console.print(f"✅ [bold green]{provider.title()} configuration successful![/bold green]\n") if not configured_providers: console.print("⚠️ At least one LLM provider is required to continue") else: # Check if there are any remaining providers to configure # remaining_providers has already filtered out configured providers, and will add "done" and "cancel" options actual_remaining = len([p for p in remaining_providers if p.value not in ["done", "cancel"]]) if actual_remaining > 0: continue_confirm = questionary.confirm("Continue configuring other LLM providers?", default=False, instruction="(y/N)").ask() if continue_confirm is None: # User pressed Ctrl+C return None if not continue_confirm: break else: # All providers are configured, automatically enter next step break return configured_providers def collect_all_selected_models(llm_configs: List[Dict]) -> List[str]: """Collect all models from all configured providers, remove duplicates and maintain order.""" seen = set() ordered_models: List[str] = [] for conf in llm_configs or []: models_str = (conf.get("models") or "").strip() if not models_str: continue for m in models_str.split(","): model = m.strip() if model and model not in seen: seen.add(model) ordered_models.append(model) return ordered_models def select_default_model_interactively(all_models: List[str]) -> Optional[str]: """Provide interactive selection of default model using arrow keys; return None if no models or user cancels.""" if not all_models: return None # Default value: first item, but allow user to change default_choice_value = all_models[0] choices = [ questionary.Choice( title=(m if m != default_choice_value else f"{m} (default)"), value=m, shortcut_key=None, ) for m in all_models ] console.print("\n⭐ Please select the default model for the session (can be modified in .env)") selected = questionary.select( "Default model:", choices=choices, default=default_choice_value, instruction="Use arrow keys to navigate, press Enter to confirm", ).ask() if selected is None: # User pressed Ctrl+C return None return selected or default_choice_value