agentkit-sdk 0.3.2__tar.gz → 0.3.4__tar.gz

{agentkit_sdk-0.3.2 → agentkit_sdk-0.3.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: agentkit-sdk
-Version: 0.3.2
+Version: 0.3.4
 Summary: Developer SDK for building personalized voice AI assistants
 Requires-Python: >=3.11
 Description-Content-Type: text/markdown

{agentkit_sdk-0.3.2 → agentkit_sdk-0.3.4}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "agentkit-sdk"
-version = "0.3.2"
+version = "0.3.4"
 description = "Developer SDK for building personalized voice AI assistants"
 readme = "README.md"
 requires-python = ">=3.11"

{agentkit_sdk-0.3.2 → agentkit_sdk-0.3.4}/src/agentkit/__init__.py

@@ -1,3 +1,3 @@
 """AgentKit - Developer SDK for building personalized voice AI assistants"""
 
-__version__ = "0.3.2"
+__version__ = "0.3.4"

{agentkit_sdk-0.3.2 → agentkit_sdk-0.3.4}/src/agentkit/cli/init.py

@@ -16,6 +16,7 @@ BUILTIN_PROVIDERS = {
     },
     "tts": {
         "sarvam":     {"desc": "Sarvam AI",   "env": "SARVAM_API_KEY",     "defaults": {"voice": "meera", "language_code": "hi-IN"}},
+        "deepgram":   {"desc": "Deepgram Aura", "env": "DEEPGRAM_API_KEY",  "defaults": {"model": "aura-asteria-en"}},
         "elevenlabs": {"desc": "ElevenLabs",  "env": "ELEVENLABS_API_KEY", "defaults": {"voice_id": "pNInz6obpgDQGcFmaJgB"}},
     },
     "llm": {

{agentkit_sdk-0.3.2 → agentkit_sdk-0.3.4}/src/agentkit/core/pipeline.py

@@ -1,4 +1,5 @@
 import asyncio
+import time
 from dataclasses import dataclass, field
 from typing import AsyncIterator
 
@@ -76,7 +77,7 @@ class VoicePipeline:
                 yield audio_chunk
         
         self.conversation_history.append(Message(role="assistant", content=full_response))
-        self.turns.append(Turn(user_message=transcription, assistant_message=full_response))
+        self.turns.append(Turn(user_message=transcription, assistant_message=full_response, timestamp=time.time()))
         self._last_response = full_response
 
     async def process_text(
@@ -112,7 +113,7 @@ class VoicePipeline:
                 yield audio_chunk
         
         self.conversation_history.append(Message(role="assistant", content=full_response))
-        self.turns.append(Turn(user_message=user_message, assistant_message=full_response))
+        self.turns.append(Turn(user_message=user_message, assistant_message=full_response, timestamp=time.time()))
         self._last_response = full_response
 
     def _is_sentence_end(self, text: str) -> bool:

{agentkit_sdk-0.3.2 → agentkit_sdk-0.3.4}/src/agentkit/providers/__init__.py

@@ -198,6 +198,7 @@ def _register_builtins():
     from agentkit.providers.llm.groq import GroqLLM
     from agentkit.providers.llm.openai import OpenAILLM
     from agentkit.providers.tts.sarvam import SarvamTTS
+    from agentkit.providers.tts.deepgram import DeepgramTTS
     from agentkit.providers.tts.elevenlabs import ElevenLabsTTS
     from agentkit.memory.markdown import MarkdownMemory
     from agentkit.memory.vector import VectorMemory
@@ -210,6 +211,7 @@ def _register_builtins():
     registry.register("llm", "openai", OpenAILLM)
 
     registry.register("tts", "sarvam", SarvamTTS)
+    registry.register("tts", "deepgram", DeepgramTTS)
     registry.register("tts", "elevenlabs", ElevenLabsTTS)
 
     registry.register("memory", "markdown", MarkdownMemory)

agentkit_sdk-0.3.4/src/agentkit/providers/tts/__init__.py

@@ -0,0 +1,6 @@
+from .base import BaseTTS
+from .deepgram import DeepgramTTS
+from .elevenlabs import ElevenLabsTTS
+from .sarvam import SarvamTTS
+
+__all__ = ["BaseTTS", "DeepgramTTS", "ElevenLabsTTS", "SarvamTTS"]

agentkit_sdk-0.3.4/src/agentkit/providers/tts/deepgram.py

@@ -0,0 +1,42 @@
+from typing import AsyncIterator
+
+import httpx
+
+from .base import BaseTTS
+
+
+class DeepgramTTS(BaseTTS):
+    """Deepgram Aura text-to-speech provider."""
+
+    def __init__(self, api_key: str, model: str = "aura-asteria-en", encoding: str = "linear16", sample_rate: int = 16000):
+        self.api_key = api_key
+        self.model = model
+        self.encoding = encoding
+        self.sample_rate = sample_rate
+        self.base_url = "https://api.deepgram.com/v1/speak"
+
+    async def synthesize_stream(self, text: str) -> AsyncIterator[bytes]:
+        async with httpx.AsyncClient(timeout=30.0) as client:
+            response = await client.post(
+                self.base_url,
+                json={"text": text},
+                headers={
+                    "Authorization": f"Token {self.api_key}",
+                    "Content-Type": "application/json",
+                },
+                params={
+                    "model": self.model,
+                    "encoding": self.encoding,
+                    "sample_rate": self.sample_rate,
+                },
+            )
+            if response.status_code == 200:
+                yield response.content
+
+    async def synthesize(self, text: str) -> bytes:
+        async for chunk in self.synthesize_stream(text):
+            return chunk
+        return b""
+
+    async def close(self):
+        pass

{agentkit_sdk-0.3.2 → agentkit_sdk-0.3.4}/src/agentkit_sdk.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: agentkit-sdk
-Version: 0.3.2
+Version: 0.3.4
 Summary: Developer SDK for building personalized voice AI assistants
 Requires-Python: >=3.11
 Description-Content-Type: text/markdown

{agentkit_sdk-0.3.2 → agentkit_sdk-0.3.4}/src/agentkit_sdk.egg-info/SOURCES.txt

@@ -32,6 +32,7 @@ src/agentkit/providers/stt/deepgram.py
 src/agentkit/providers/stt/sarvam.py
 src/agentkit/providers/tts/__init__.py
 src/agentkit/providers/tts/base.py
+src/agentkit/providers/tts/deepgram.py
 src/agentkit/providers/tts/elevenlabs.py
 src/agentkit/providers/tts/sarvam.py
 src/agentkit/server/__init__.py

agentkit_sdk-0.3.2/src/agentkit/providers/tts/__init__.py

@@ -1,5 +0,0 @@
-from .base import BaseTTS
-from .sarvam import SarvamTTS
-from .elevenlabs import ElevenLabsTTS
-
-__all__ = ["BaseTTS", "SarvamTTS", "ElevenLabsTTS"]