apiVersion: capsule.dev/v0.1 kind: Capsule name: cverse-voice-llm-inference-adapter version: 0.1.0 type: adapter purpose: summary: 'Provides adapters for voice-enabled Large Language Models, including specific integrations for Doubao and Qwen Omni, and a Persona Agent framework for complex conversational flows. This allows the core inference service to handle real-time voice interactions. ' owns: - Base VoiceLLM plugin interface - Doubao VoiceLLM integration (protocol, config, real-time client) - Qwen Omni real-time integration - Persona Agent framework (supervisor, subagents, tools, schemas) does_not_own: - The core inference service - Other AI model types (ASR, LLM, TTS, Avatar, etc.) interfaces: provides: - kind: library name: voice-llm-plugin description: VoiceLLM implementations conforming to the inference core's plugin interface. requires: - kind: library name: llm-plugin from_capsule: cverse-llm-inference-adapter description: Uses LLM plugins for text-based reasoning within the Persona Agent. dependencies: capsules: - name: cverse-inference-core version: '>=0.1.0' - name: cverse-llm-inference-adapter version: '>=0.1.0' agent: summary_for_ai: 'An AI agent working on this capsule would focus on integrating new real-time conversational AI models, enhancing the Persona Agent''s capabilities (e.g., adding new tools, improving agent reasoning), or optimizing real-time interaction flows. It requires knowledge of conversational AI, agent frameworks, and Python plugin development. ' avoid: - Modifying the core inference service or other AI model types. verification: invariants: - VoiceLLM plugins must process real-time audio/text inputs and generate appropriate responses. - Persona Agent must maintain conversational context and execute tasks reliably. - Plugins must adhere to the defined VoiceLLM plugin interface. x-reuse: notes: 'The `doubao_config.py` contains specific Doubao API configurations. The Persona Agent''s default tools and agent logic might need customization for different use cases or domains. ' x-reconstruct: install: install.json