Phase 5: AI provider system with local and cloud support
- Implement AIProvider base interface with chat() and is_available() - Add LocalProvider connecting to bundled llama-server via OpenAI SDK - Add OpenAIProvider for direct OpenAI API access - Add AnthropicProvider for Anthropic Claude API - Add LiteLLMProvider for multi-provider gateway - Build AIProviderService with provider routing, auto-selection, and transcript context injection - Add ai.chat IPC handler supporting chat, list_providers, set_provider, and configure actions - Add ai_chat, ai_list_providers, ai_configure Tauri commands - Build interactive AIChatPanel with message history, quick actions (Summarize, Action Items), and transcript context awareness - Tests: 30 Python, 6 Rust, 0 Svelte errors Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
@@ -135,6 +135,83 @@ def make_export_handler() -> HandlerFunc:
|
||||
return handler
|
||||
|
||||
|
||||
def make_ai_chat_handler() -> HandlerFunc:
|
||||
"""Create an AI chat handler with persistent AIProviderService."""
|
||||
from voice_to_notes.services.ai_provider import create_default_service
|
||||
|
||||
service = create_default_service()
|
||||
|
||||
def handler(msg: IPCMessage) -> IPCMessage:
|
||||
payload = msg.payload
|
||||
action = payload.get("action", "chat")
|
||||
|
||||
if action == "list_providers":
|
||||
return IPCMessage(
|
||||
id=msg.id,
|
||||
type="ai.providers",
|
||||
payload={"providers": service.list_providers()},
|
||||
)
|
||||
|
||||
if action == "set_provider":
|
||||
service.set_active(payload["provider"])
|
||||
return IPCMessage(
|
||||
id=msg.id,
|
||||
type="ai.provider_set",
|
||||
payload={"provider": payload["provider"]},
|
||||
)
|
||||
|
||||
if action == "configure":
|
||||
# Re-create a provider with custom settings
|
||||
provider_name = payload.get("provider", "")
|
||||
config = payload.get("config", {})
|
||||
if provider_name == "local":
|
||||
from voice_to_notes.providers.local_provider import LocalProvider
|
||||
|
||||
service.register_provider("local", LocalProvider(
|
||||
base_url=config.get("base_url", "http://localhost:8080"),
|
||||
model=config.get("model", "local"),
|
||||
))
|
||||
elif provider_name == "openai":
|
||||
from voice_to_notes.providers.openai_provider import OpenAIProvider
|
||||
|
||||
service.register_provider("openai", OpenAIProvider(
|
||||
api_key=config.get("api_key"),
|
||||
model=config.get("model", "gpt-4o-mini"),
|
||||
))
|
||||
elif provider_name == "anthropic":
|
||||
from voice_to_notes.providers.anthropic_provider import AnthropicProvider
|
||||
|
||||
service.register_provider("anthropic", AnthropicProvider(
|
||||
api_key=config.get("api_key"),
|
||||
model=config.get("model", "claude-sonnet-4-6"),
|
||||
))
|
||||
elif provider_name == "litellm":
|
||||
from voice_to_notes.providers.litellm_provider import LiteLLMProvider
|
||||
|
||||
service.register_provider("litellm", LiteLLMProvider(
|
||||
model=config.get("model", "gpt-4o-mini"),
|
||||
))
|
||||
return IPCMessage(
|
||||
id=msg.id,
|
||||
type="ai.configured",
|
||||
payload={"provider": provider_name},
|
||||
)
|
||||
|
||||
# Default: chat
|
||||
response = service.chat(
|
||||
messages=payload.get("messages", []),
|
||||
transcript_context=payload.get("transcript_context", ""),
|
||||
**{k: v for k, v in payload.items() if k not in ("action", "messages", "transcript_context")},
|
||||
)
|
||||
return IPCMessage(
|
||||
id=msg.id,
|
||||
type="ai.response",
|
||||
payload={"response": response},
|
||||
)
|
||||
|
||||
return handler
|
||||
|
||||
|
||||
def hardware_detect_handler(msg: IPCMessage) -> IPCMessage:
|
||||
"""Detect hardware capabilities and return recommendations."""
|
||||
from voice_to_notes.hardware.detect import detect_hardware
|
||||
|
||||
@@ -8,6 +8,7 @@ import sys
|
||||
from voice_to_notes.ipc.handlers import (
|
||||
HandlerRegistry,
|
||||
hardware_detect_handler,
|
||||
make_ai_chat_handler,
|
||||
make_diarize_handler,
|
||||
make_export_handler,
|
||||
make_pipeline_handler,
|
||||
@@ -27,6 +28,7 @@ def create_registry() -> HandlerRegistry:
|
||||
registry.register("diarize.start", make_diarize_handler())
|
||||
registry.register("pipeline.start", make_pipeline_handler())
|
||||
registry.register("export.start", make_export_handler())
|
||||
registry.register("ai.chat", make_ai_chat_handler())
|
||||
return registry
|
||||
|
||||
|
||||
|
||||
@@ -2,4 +2,68 @@
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
# TODO: Implement Anthropic provider
|
||||
import os
|
||||
from typing import Any
|
||||
|
||||
from voice_to_notes.providers.base import AIProvider
|
||||
|
||||
|
||||
class AnthropicProvider(AIProvider):
|
||||
"""Connects to the Anthropic API."""
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
api_key: str | None = None,
|
||||
model: str = "claude-sonnet-4-6",
|
||||
) -> None:
|
||||
self._api_key = api_key or os.environ.get("ANTHROPIC_API_KEY", "")
|
||||
self._model = model
|
||||
self._client: Any = None
|
||||
|
||||
def _ensure_client(self) -> Any:
|
||||
if self._client is not None:
|
||||
return self._client
|
||||
|
||||
if not self._api_key:
|
||||
raise RuntimeError(
|
||||
"Anthropic API key not configured. Set ANTHROPIC_API_KEY or provide it in settings."
|
||||
)
|
||||
|
||||
try:
|
||||
import anthropic
|
||||
|
||||
self._client = anthropic.Anthropic(api_key=self._api_key)
|
||||
except ImportError:
|
||||
raise RuntimeError("anthropic package is required. Install with: pip install anthropic")
|
||||
return self._client
|
||||
|
||||
def chat(self, messages: list[dict[str, str]], **kwargs: Any) -> str:
|
||||
client = self._ensure_client()
|
||||
|
||||
# Anthropic expects a system message separately
|
||||
system_msg = ""
|
||||
chat_messages = []
|
||||
for msg in messages:
|
||||
if msg.get("role") == "system":
|
||||
system_msg = msg.get("content", "")
|
||||
else:
|
||||
chat_messages.append(msg)
|
||||
|
||||
create_kwargs: dict[str, Any] = {
|
||||
"model": kwargs.get("model", self._model),
|
||||
"messages": chat_messages,
|
||||
"max_tokens": kwargs.get("max_tokens", 2048),
|
||||
}
|
||||
if system_msg:
|
||||
create_kwargs["system"] = system_msg
|
||||
|
||||
response = client.messages.create(**create_kwargs)
|
||||
# Anthropic returns content blocks
|
||||
return "".join(block.text for block in response.content if hasattr(block, "text"))
|
||||
|
||||
def is_available(self) -> bool:
|
||||
return bool(self._api_key)
|
||||
|
||||
@property
|
||||
def name(self) -> str:
|
||||
return "Anthropic"
|
||||
|
||||
@@ -3,7 +3,6 @@
|
||||
from __future__ import annotations
|
||||
|
||||
from abc import ABC, abstractmethod
|
||||
from collections.abc import AsyncIterator
|
||||
from typing import Any
|
||||
|
||||
|
||||
@@ -11,13 +10,17 @@ class AIProvider(ABC):
|
||||
"""Base interface for all AI providers."""
|
||||
|
||||
@abstractmethod
|
||||
async def chat(self, messages: list[dict[str, Any]], config: dict[str, Any]) -> str:
|
||||
"""Send a chat completion request and return the response."""
|
||||
def chat(self, messages: list[dict[str, str]], **kwargs: Any) -> str:
|
||||
"""Send a chat completion request and return the full response text."""
|
||||
...
|
||||
|
||||
@abstractmethod
|
||||
async def stream(
|
||||
self, messages: list[dict[str, Any]], config: dict[str, Any]
|
||||
) -> AsyncIterator[str]:
|
||||
"""Send a streaming chat request, yielding tokens as they arrive."""
|
||||
def is_available(self) -> bool:
|
||||
"""Check if this provider is configured and available."""
|
||||
...
|
||||
|
||||
@property
|
||||
@abstractmethod
|
||||
def name(self) -> str:
|
||||
"""Provider display name."""
|
||||
...
|
||||
|
||||
@@ -2,4 +2,41 @@
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
# TODO: Implement LiteLLM provider
|
||||
from typing import Any
|
||||
|
||||
from voice_to_notes.providers.base import AIProvider
|
||||
|
||||
|
||||
class LiteLLMProvider(AIProvider):
|
||||
"""Routes through LiteLLM for access to 100+ LLM providers."""
|
||||
|
||||
def __init__(self, model: str = "gpt-4o-mini", **kwargs: Any) -> None:
|
||||
self._model = model
|
||||
self._extra_kwargs = kwargs
|
||||
|
||||
def chat(self, messages: list[dict[str, str]], **kwargs: Any) -> str:
|
||||
try:
|
||||
import litellm
|
||||
except ImportError:
|
||||
raise RuntimeError("litellm package is required. Install with: pip install litellm")
|
||||
|
||||
merged_kwargs = {**self._extra_kwargs, **kwargs}
|
||||
response = litellm.completion(
|
||||
model=merged_kwargs.get("model", self._model),
|
||||
messages=messages,
|
||||
temperature=merged_kwargs.get("temperature", 0.7),
|
||||
max_tokens=merged_kwargs.get("max_tokens", 2048),
|
||||
)
|
||||
return response.choices[0].message.content or ""
|
||||
|
||||
def is_available(self) -> bool:
|
||||
try:
|
||||
import litellm # noqa: F401
|
||||
|
||||
return True
|
||||
except ImportError:
|
||||
return False
|
||||
|
||||
@property
|
||||
def name(self) -> str:
|
||||
return "LiteLLM"
|
||||
|
||||
@@ -2,8 +2,57 @@
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
import sys
|
||||
from typing import Any
|
||||
|
||||
# TODO: Implement local provider
|
||||
# - Connect to llama-server on localhost:{port}
|
||||
# - Use openai SDK with custom base_url
|
||||
# - Support chat and streaming
|
||||
from voice_to_notes.providers.base import AIProvider
|
||||
|
||||
|
||||
class LocalProvider(AIProvider):
|
||||
"""Connects to bundled llama-server via its OpenAI-compatible API."""
|
||||
|
||||
def __init__(self, base_url: str = "http://localhost:8080", model: str = "local") -> None:
|
||||
self._base_url = base_url.rstrip("/")
|
||||
self._model = model
|
||||
self._client: Any = None
|
||||
|
||||
def _ensure_client(self) -> Any:
|
||||
if self._client is not None:
|
||||
return self._client
|
||||
|
||||
try:
|
||||
from openai import OpenAI
|
||||
|
||||
self._client = OpenAI(
|
||||
base_url=f"{self._base_url}/v1",
|
||||
api_key="not-needed", # llama-server doesn't require an API key
|
||||
)
|
||||
except ImportError:
|
||||
raise RuntimeError(
|
||||
"openai package is required for local AI. Install with: pip install openai"
|
||||
)
|
||||
return self._client
|
||||
|
||||
def chat(self, messages: list[dict[str, str]], **kwargs: Any) -> str:
|
||||
client = self._ensure_client()
|
||||
response = client.chat.completions.create(
|
||||
model=self._model,
|
||||
messages=messages,
|
||||
temperature=kwargs.get("temperature", 0.7),
|
||||
max_tokens=kwargs.get("max_tokens", 2048),
|
||||
)
|
||||
return response.choices[0].message.content or ""
|
||||
|
||||
def is_available(self) -> bool:
|
||||
try:
|
||||
import urllib.request
|
||||
|
||||
req = urllib.request.Request(f"{self._base_url}/health", method="GET")
|
||||
with urllib.request.urlopen(req, timeout=2) as resp:
|
||||
return resp.status == 200
|
||||
except Exception:
|
||||
return False
|
||||
|
||||
@property
|
||||
def name(self) -> str:
|
||||
return "Local (llama-server)"
|
||||
|
||||
@@ -2,4 +2,52 @@
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
# TODO: Implement OpenAI provider
|
||||
import os
|
||||
from typing import Any
|
||||
|
||||
from voice_to_notes.providers.base import AIProvider
|
||||
|
||||
|
||||
class OpenAIProvider(AIProvider):
|
||||
"""Connects to the OpenAI API."""
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
api_key: str | None = None,
|
||||
model: str = "gpt-4o-mini",
|
||||
) -> None:
|
||||
self._api_key = api_key or os.environ.get("OPENAI_API_KEY", "")
|
||||
self._model = model
|
||||
self._client: Any = None
|
||||
|
||||
def _ensure_client(self) -> Any:
|
||||
if self._client is not None:
|
||||
return self._client
|
||||
|
||||
if not self._api_key:
|
||||
raise RuntimeError("OpenAI API key not configured. Set OPENAI_API_KEY or provide it in settings.")
|
||||
|
||||
try:
|
||||
from openai import OpenAI
|
||||
|
||||
self._client = OpenAI(api_key=self._api_key)
|
||||
except ImportError:
|
||||
raise RuntimeError("openai package is required. Install with: pip install openai")
|
||||
return self._client
|
||||
|
||||
def chat(self, messages: list[dict[str, str]], **kwargs: Any) -> str:
|
||||
client = self._ensure_client()
|
||||
response = client.chat.completions.create(
|
||||
model=kwargs.get("model", self._model),
|
||||
messages=messages,
|
||||
temperature=kwargs.get("temperature", 0.7),
|
||||
max_tokens=kwargs.get("max_tokens", 2048),
|
||||
)
|
||||
return response.choices[0].message.content or ""
|
||||
|
||||
def is_available(self) -> bool:
|
||||
return bool(self._api_key)
|
||||
|
||||
@property
|
||||
def name(self) -> str:
|
||||
return "OpenAI"
|
||||
|
||||
@@ -2,12 +2,103 @@
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
import sys
|
||||
from typing import Any
|
||||
|
||||
from voice_to_notes.providers.base import AIProvider
|
||||
|
||||
|
||||
class AIProviderService:
|
||||
"""Manages AI provider selection and routes chat/summarize requests."""
|
||||
"""Manages AI provider selection and routes chat requests."""
|
||||
|
||||
# TODO: Implement provider routing
|
||||
# - Select provider based on config (local, openai, anthropic, litellm)
|
||||
# - Forward chat messages
|
||||
# - Handle streaming responses
|
||||
pass
|
||||
def __init__(self) -> None:
|
||||
self._providers: dict[str, AIProvider] = {}
|
||||
self._active_provider: str | None = None
|
||||
|
||||
def register_provider(self, name: str, provider: AIProvider) -> None:
|
||||
"""Register an AI provider."""
|
||||
self._providers[name] = provider
|
||||
|
||||
def set_active(self, name: str) -> None:
|
||||
"""Set the active provider by name."""
|
||||
if name not in self._providers:
|
||||
raise ValueError(f"Unknown provider: {name}. Available: {list(self._providers.keys())}")
|
||||
self._active_provider = name
|
||||
|
||||
def get_active(self) -> AIProvider | None:
|
||||
"""Get the currently active provider."""
|
||||
if self._active_provider:
|
||||
return self._providers.get(self._active_provider)
|
||||
# Auto-select first available provider
|
||||
for name, provider in self._providers.items():
|
||||
if provider.is_available():
|
||||
self._active_provider = name
|
||||
return provider
|
||||
return None
|
||||
|
||||
def list_providers(self) -> list[dict[str, Any]]:
|
||||
"""List all registered providers with their status."""
|
||||
return [
|
||||
{
|
||||
"name": name,
|
||||
"display_name": provider.name,
|
||||
"available": provider.is_available(),
|
||||
"active": name == self._active_provider,
|
||||
}
|
||||
for name, provider in self._providers.items()
|
||||
]
|
||||
|
||||
def chat(
|
||||
self,
|
||||
messages: list[dict[str, str]],
|
||||
transcript_context: str = "",
|
||||
**kwargs: Any,
|
||||
) -> str:
|
||||
"""Send a chat request to the active provider.
|
||||
|
||||
Automatically prepends transcript context as a system message if provided.
|
||||
"""
|
||||
provider = self.get_active()
|
||||
if provider is None:
|
||||
raise RuntimeError(
|
||||
"No AI provider available. Configure a provider in settings or start the local AI server."
|
||||
)
|
||||
|
||||
# Build messages with transcript context
|
||||
full_messages: list[dict[str, str]] = []
|
||||
if transcript_context:
|
||||
full_messages.append({
|
||||
"role": "system",
|
||||
"content": (
|
||||
"You are a helpful assistant analyzing a transcript. "
|
||||
"Here is the transcript for context:\n\n"
|
||||
f"{transcript_context}\n\n"
|
||||
"Answer the user's questions about this transcript. "
|
||||
"Be concise and helpful."
|
||||
),
|
||||
})
|
||||
|
||||
full_messages.extend(messages)
|
||||
|
||||
print(
|
||||
f"[sidecar] AI chat via {provider.name}, {len(full_messages)} messages",
|
||||
file=sys.stderr,
|
||||
flush=True,
|
||||
)
|
||||
|
||||
return provider.chat(full_messages, **kwargs)
|
||||
|
||||
|
||||
def create_default_service() -> AIProviderService:
|
||||
"""Create an AIProviderService with all supported providers registered."""
|
||||
from voice_to_notes.providers.anthropic_provider import AnthropicProvider
|
||||
from voice_to_notes.providers.litellm_provider import LiteLLMProvider
|
||||
from voice_to_notes.providers.local_provider import LocalProvider
|
||||
from voice_to_notes.providers.openai_provider import OpenAIProvider
|
||||
|
||||
service = AIProviderService()
|
||||
service.register_provider("local", LocalProvider())
|
||||
service.register_provider("openai", OpenAIProvider())
|
||||
service.register_provider("anthropic", AnthropicProvider())
|
||||
service.register_provider("litellm", LiteLLMProvider())
|
||||
return service
|
||||
|
||||
Reference in New Issue
Block a user