Claude/multimodal ye linghua refactor 01 p tmtnewy3wnkopm qjd bjyv#16
Closed
win10ogod wants to merge 4 commits intoMiniMax-AI:mainfrom
Closed
Conversation
…c & OpenAI) Major Changes: 1. **Project Rename**: mini-agent -> ye_linghua (叶灵华) - Renamed package: mini_agent -> ye_linghua - Renamed CLI command: mini-agent -> ye-linghua - Updated all imports and references - Added YeLinghua alias for Agent class 2. **OpenAI Multimodal Support** - Enhanced OpenAI client with multimodal input support - Support text + image (URL or base64) messages - Updated documentation 3. **YAML-based Personality System** - Created personality.yaml for character configuration - Created prompts.yaml for prompt templates - Implemented PersonalityLoader for dynamic prompt generation - Backward compatible with legacy system_prompt.md 4. **Web Configuration Interface** - FastAPI-based web UI for editing personality and prompts - Real-time preview of generated system prompts - YAML validation and syntax highlighting - New CLI command: ye-linghua-config 5. **Default Personality: Ye Linghua (叶灵华)** - Programming-loving AI girl character - Friendly, enthusiastic, and professional - Customizable traits, interests, and behaviors - Chinese + English bilingual support 6. **Updated Configuration** - Added agent.use_personality config option - Added agent.personality_path and prompts_path - Updated config-example.yaml with new options - All config paths now use ye-linghua prefix 7. **Documentation** - Created comprehensive README_ZH.md - Updated all examples with new imports - Added Web UI usage guide - Migration guide from Mini-Agent Breaking Changes: - Package name changed: mini_agent -> ye_linghua - CLI command changed: mini-agent -> ye-linghua - Config directory changed: ~/.mini-agent -> ~/.ye-linghua Migration: - Update imports: from mini_agent -> from ye_linghua - Use YeLinghua or Agent (both work) - Legacy configs still supported for backward compatibility
…port Implement a modular plugin architecture supporting: - Python-based plugins with base classes (Plugin, ReplyHandler) - PowerShell/Shell script plugins with JSON communication - Timer plugin for scheduled reminders with tag-based interface (<set-timer>, <list-timers>, <remove-timer>) - Notification plugin with cross-platform desktop notifications (Windows/Linux/macOS) - PluginLoader with hot-reload capability - PluginRegistry for centralized management - ReplyHandler chain for processing LLM responses by priority - Context and prompt injection into LLM system prompts - Auto-discovery from plugins directory Configuration integration: - Added PluginsConfig to config system - Updated config-example.yaml with plugin settings - Support for enabling/disabling individual plugins Documentation: - Comprehensive PLUGINS.md with usage examples - Architecture explanations and best practices
Agent Integration: - Add plugin_registry, platform, user_id, session_id parameters to Agent.__init__() - Implement _build_plugin_context() to convert agent state to PluginContext - Implement _get_plugin_prompt_extensions() to inject plugin prompts into system prompt - Implement _process_reply_with_plugins() to process LLM responses through ReplyHandler chain - Update run() method to process responses with plugin handlers before display - Plugin prompt extensions are automatically injected into system prompt on initialization CLI Integration: - Initialize PluginRegistry and PluginLoader in run_agent() - Load built-in plugins (Timer & Notification) based on config - Auto-discover and load plugins from plugins directory - Pass plugin_registry to Agent constructor - Generate unique session_id for each session - Display active plugins count in session info - Properly shutdown all plugins on exit Features: - Plugin prompt extensions visible to LLM in system prompt - ReplyHandler chain processes responses by priority - Tag-based functionality (<set-timer>, <notify>) works automatically - Platform context (cli) passed to plugins - Session tracking with unique IDs
- Document automatic plugin integration in Agent - Explain system prompt injection and ReplyHandler chain - Provide CLI usage examples with plugin loading output - Add programmatic usage example for custom integrations - Show how plugins work transparently with Agent runtime
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
No description provided.