feature: unify model metadata management - MCP server

[JoeGaffney]

No description provided.

[Copilot]

Pull Request Overview

This PR refactors the schema system to unify model metadata management across images, texts, and videos. Key changes include:

• Consolidation of separate local/external model metadata into unified MODEL_META dictionaries
• Replacement of message-based text API with simpler prompt/system_prompt fields
• Addition of model capability validation using @model_validator decorators
• New model listing endpoints for all resource types
• Implementation of MCP server and agentic layer with Gradio UI
• Docker Compose updates to add MCP and agentic services

Reviewed Changes

Copilot reviewed 47 out of 47 changed files in this pull request and generated 10 comments.

Show a summary per file

File	Description
workers/videos/schemas.py	Unified video model metadata with capability validation and inferred mode detection
workers/texts/schemas.py	Simplified text API interface (prompt/system_prompt instead of messages)
workers/images/schemas.py	Added capability validation and inferred mode detection for images
workers/texts/models/qwen.py	Updated to use new prompt/system_prompt structure
workers/texts/external_models/openai.py	Updated to use new text request structure
api/*/router.py	Added model listing endpoints for images, texts, and videos
docker-compose.yml	Added mcp and agentic services
agentic/*	New agentic orchestration layer with Gradio UI

Comments suppressed due to low confidence (1)

workers/texts/external_models/openai.py:13

• GPT-5 does not exist as of January 2025. This model mapping references a non-existent model version. Consider removing this entry or updating it to a valid OpenAI model.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.