solana-agent 22.0.1__py3-none-any.whl → 22.0.3__py3-none-any.whl

solana_agent/adapters/llm_adapter.py

@@ -22,12 +22,12 @@ class OpenAIAdapter(LLMProvider):
         self.text_model = "gpt-4o-mini"
         self.internet_search_model = "gpt-4o-mini-search-preview"
         self.transcription_model = "gpt-4o-mini-transcribe"
-        self.tts_model = "tts-1"
+        self.tts_model = "gpt-4o-mini-tts"
 
     async def tts(
         self,
         text: str,
-        instructions: str = "",
+        instructions: str = "You speak in a friendly and helpful manner.",
         voice: Literal["alloy", "ash", "ballad", "coral", "echo",
                        "fable", "onyx", "nova", "sage", "shimmer"] = "nova",
         response_format: Literal['mp3', 'opus',
@@ -45,16 +45,16 @@ class OpenAIAdapter(LLMProvider):
             Audio bytes as they become available
         """
         try:
-            stream = self.client.audio.speech.create(
+            with self.client.audio.speech.with_streaming_response.create(
                 model=self.tts_model,
                 voice=voice,
+                instructions=instructions,
                 input=text,
                 response_format=response_format
-            )
-
-            # Stream the bytes in chunks
-            for chunk in stream.iter_bytes(chunk_size=1024 * 16):  # 16KB chunks
-                yield chunk
+            ) as stream:
+                # Stream the bytes in 16KB chunks
+                for chunk in stream.iter_bytes(chunk_size=1024 * 16):
+                    yield chunk
 
         except Exception as e:
             print(f"Error in text_to_speech: {str(e)}")
@@ -66,7 +66,7 @@ class OpenAIAdapter(LLMProvider):
             print(f"Error in text_to_speech: {str(e)}")
             import traceback
             print(traceback.format_exc())
-            yield f"I apologize, but I encountered an error converting text to speech: {str(e)}"
+            yield b""  # Return empty bytes on error
 
     async def transcribe_audio(
         self,
@@ -85,16 +85,14 @@ class OpenAIAdapter(LLMProvider):
             Transcript text chunks as they become available
         """
         try:
-            stream = self.client.audio.transcriptions.create(
+            with self.client.audio.transcriptions.with_streaming_response.create(
                 model=self.transcription_model,
                 file=(f"file.{input_format}", audio_bytes),
                 response_format="text",
-                stream=True
-            )
-
-            for event in stream:
-                if hasattr(event, 'text') and event.text:
-                    yield event.text
+            ) as stream:
+                # Stream the text in 16KB chunks
+                for chunk in stream.iter_text(chunk_size=1024 * 16):
+                    yield chunk
 
         except Exception as e:
             print(f"Error in transcribe_audio: {str(e)}")

solana_agent/client/solana_agent.py

@@ -49,7 +49,7 @@ class SolanaAgent(SolanaAgentInterface):
         output_format: Literal["text", "audio"] = "text",
         audio_voice: Literal["alloy", "ash", "ballad", "coral", "echo",
                              "fable", "onyx", "nova", "sage", "shimmer"] = "nova",
-        audio_instructions: Optional[str] = None,
+        audio_instructions: str = "You speak in a friendly and helpful manner.",
         audio_output_format: Literal['mp3', 'opus',
                                      'aac', 'flac', 'wav', 'pcm'] = "aac",
         audio_input_format: Literal[
@@ -66,7 +66,7 @@ class SolanaAgent(SolanaAgentInterface):
             prompt: Optional prompt for the agent
             output_format: Response format ("text" or "audio")
             audio_voice: Voice to use for audio output
-            audio_instructions: Not used currently
+            audio_instructions: Audio voice instructions
             audio_output_format: Audio output format
             audio_input_format: Audio input format
             router: Optional routing service for processing

solana_agent/interfaces/client/client.py

@@ -17,7 +17,7 @@ class SolanaAgent(ABC):
         output_format: Literal["text", "audio"] = "text",
         audio_voice: Literal["alloy", "ash", "ballad", "coral", "echo",
                              "fable", "onyx", "nova", "sage", "shimmer"] = "nova",
-        audio_instructions: Optional[str] = None,
+        audio_instructions: str = "You speak in a friendly and helpful manner.",
         audio_output_format: Literal['mp3', 'opus',
                                      'aac', 'flac', 'wav', 'pcm'] = "aac",
         audio_input_format: Literal[

solana_agent/interfaces/providers/llm.py

@@ -31,7 +31,7 @@ class LLMProvider(ABC):
     async def tts(
         self,
         text: str,
-        instructions: str = "",
+        instructions: str = "You speak in a friendly and helpful manner.",
         voice: Literal["alloy", "ash", "ballad", "coral", "echo",
                        "fable", "onyx", "nova", "sage", "shimmer"] = "nova",
         response_format: Literal['mp3', 'opus',

solana_agent/interfaces/services/agent.py

@@ -27,7 +27,7 @@ class AgentService(ABC):
         output_format: Literal["text", "audio"] = "text",
         audio_voice: Literal["alloy", "ash", "ballad", "coral", "echo",
                              "fable", "onyx", "nova", "sage", "shimmer"] = "nova",
-        audio_instructions: Optional[str] = None,
+        audio_instructions: str = "You speak in a friendly and helpful manner.",
         audio_output_format: Literal['mp3', 'opus',
                                      'aac', 'flac', 'wav', 'pcm'] = "aac",
         audio_input_format: Literal[

solana_agent/interfaces/services/query.py

@@ -13,7 +13,7 @@ class QueryService(ABC):
         output_format: Literal["text", "audio"] = "text",
         audio_voice: Literal["alloy", "ash", "ballad", "coral", "echo",
                              "fable", "onyx", "nova", "sage", "shimmer"] = "nova",
-        audio_instructions: Optional[str] = None,
+        audio_instructions: str = "You speak in a friendly and helpful manner.",
         audio_output_format: Literal['mp3', 'opus',
                                      'aac', 'flac', 'wav', 'pcm'] = "aac",
         audio_input_format: Literal[

solana_agent/services/agent.py

@@ -169,7 +169,7 @@ class AgentService(AgentServiceInterface):
         output_format: Literal["text", "audio"] = "text",
         audio_voice: Literal["alloy", "ash", "ballad", "coral", "echo",
                              "fable", "onyx", "nova", "sage", "shimmer"] = "nova",
-        audio_instructions: Optional[str] = None,
+        audio_instructions: str = "You speak in a friendly and helpful manner.",
         audio_output_format: Literal['mp3', 'opus',
                                      'aac', 'flac', 'wav', 'pcm'] = "aac",
         audio_input_format: Literal[
@@ -276,7 +276,7 @@ class AgentService(AgentServiceInterface):
                                     tool_response += processed_chunk
 
                                 # Add to our complete text record and full audio buffer
-                                tool_response = self._remove_markdown(
+                                tool_response = self._clean_for_audio(
                                     tool_response)
                                 complete_text_response += tool_response
                                 full_response_buffer += tool_response
@@ -318,8 +318,8 @@ class AgentService(AgentServiceInterface):
 
             # For audio output, now process the complete response
             if output_format == "audio" and full_response_buffer:
-                # Clean markdown before TTS
-                full_response_buffer = self._remove_markdown(
+                # Clean text before TTS
+                full_response_buffer = self._clean_for_audio(
                     full_response_buffer)
 
                 # Process the entire response with TTS
@@ -427,14 +427,14 @@ class AgentService(AgentServiceInterface):
        - Use exact tool names as shown in AVAILABLE TOOLS
     """
 
-    def _remove_markdown(self, text: str) -> str:
-        """Remove Markdown formatting and links from text.
+    def _clean_for_audio(self, text: str) -> str:
+        """Remove Markdown formatting, emojis, and non-pronounceable characters from text.
 
         Args:
-            text: Input text with potential Markdown formatting
+            text: Input text with potential Markdown formatting and special characters
 
         Returns:
-            Clean text without Markdown formatting
+            Clean text without Markdown, emojis, and special characters
         """
         import re
 
@@ -469,4 +469,34 @@ class AgentService(AgentServiceInterface):
         # Remove multiple consecutive newlines (keep just one)
         text = re.sub(r'\n{3,}', '\n\n', text)
 
+        # Remove emojis and other non-pronounceable characters
+        # Common emoji Unicode ranges
+        emoji_pattern = re.compile(
+            "["
+            "\U0001F600-\U0001F64F"  # emoticons
+            "\U0001F300-\U0001F5FF"  # symbols & pictographs
+            "\U0001F680-\U0001F6FF"  # transport & map symbols
+            "\U0001F700-\U0001F77F"  # alchemical symbols
+            "\U0001F780-\U0001F7FF"  # Geometric Shapes
+            "\U0001F800-\U0001F8FF"  # Supplemental Arrows-C
+            "\U0001F900-\U0001F9FF"  # Supplemental Symbols and Pictographs
+            "\U0001FA00-\U0001FA6F"  # Chess Symbols
+            "\U0001FA70-\U0001FAFF"  # Symbols and Pictographs Extended-A
+            "\U00002702-\U000027B0"  # Dingbats
+            "\U000024C2-\U0000257F"  # Enclosed characters
+            "\U00002600-\U000026FF"  # Miscellaneous Symbols
+            "\U00002700-\U000027BF"  # Dingbats
+            "\U0000FE00-\U0000FE0F"  # Variation Selectors
+            "\U0001F1E0-\U0001F1FF"  # Flags (iOS)
+            "]+",
+            flags=re.UNICODE
+        )
+        text = emoji_pattern.sub(r' ', text)
+
+        # Replace special characters that can cause issues with TTS
+        text = re.sub(r'[^\w\s\.\,\;\:\?\!\'\"\-\(\)]', ' ', text)
+
+        # Replace multiple spaces with a single space
+        text = re.sub(r'\s+', ' ', text)
+
         return text.strip()

solana_agent/services/query.py

@@ -41,7 +41,7 @@ class QueryService(QueryServiceInterface):
         output_format: Literal["text", "audio"] = "text",
         audio_voice: Literal["alloy", "ash", "ballad", "coral", "echo",
                              "fable", "onyx", "nova", "sage", "shimmer"] = "nova",
-        audio_instructions: Optional[str] = None,
+        audio_instructions: str = "You speak in a friendly and helpful manner.",
         audio_output_format: Literal['mp3', 'opus',
                                      'aac', 'flac', 'wav', 'pcm'] = "aac",
         audio_input_format: Literal[
@@ -58,7 +58,7 @@ class QueryService(QueryServiceInterface):
             query: Text query or audio bytes
             output_format: Response format ("text" or "audio")
             audio_voice: Voice for TTS (text-to-speech)
-            audio_instructions: Optional instructions for TTS
+            audio_instructions: Audio voice instructions
             audio_output_format: Audio output format
             audio_input_format: Audio input format
             prompt: Optional prompt for the agent
@@ -84,7 +84,8 @@ class QueryService(QueryServiceInterface):
                     async for chunk in self.agent_service.llm_provider.tts(
                         text=response,
                         voice=audio_voice,
-                        response_format=audio_output_format
+                        response_format=audio_output_format,
+                        instructions=audio_instructions,
                     ):
                         yield chunk
                 else:

{solana_agent-22.0.1.dist-info → solana_agent-22.0.3.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: solana-agent
-Version: 22.0.1
+Version: 22.0.3
 Summary: Agentic IQ
 License: MIT
 Keywords: ai,openai,ai agents,agi
@@ -249,6 +249,15 @@ async for response in solana_agent.process("user123", "Write me a poem.", intern
     print(response, end="")
 ```
 
+### Customize Audio Voice
+
+This is an audio to audio example using the `audio_instructions` parameter.
+
+```python
+async for response in solana_agent.process("user123", audio_content, output_format="audio", audio_voice="nova", audio_input_format="webm", audio_output_format="aac", audio_instructions="You speak with an American southern accent"):
+    print(response, end="")
+```
+
 ## Tools
 
 Tools can be used from plugins like Solana Agent Kit (sakit) or via custom inline tools. Tools available via plugins integrate automatically with Solana Agent.

{solana_agent-22.0.1.dist-info → solana_agent-22.0.3.dist-info}/RECORD

@@ -1,22 +1,22 @@
 solana_agent/__init__.py,sha256=ceYeUpjIitpln8YK1r0JVJU8mzG6cRPYu-HLny3d-Tw,887
 solana_agent/adapters/__init__.py,sha256=tiEEuuy0NF3ngc_tGEcRTt71zVI58v3dYY9RvMrF2Cg,204
-solana_agent/adapters/llm_adapter.py,sha256=PsSkMrsSqZzXAL3NcQ9Zz7UCtyJYU0USvZ7uTD_I8NI,5629
+solana_agent/adapters/llm_adapter.py,sha256=ReCVQH0X0hf5NpLqEMESft5LZtPj3gDNIOBiZpClqzo,5737
 solana_agent/adapters/mongodb_adapter.py,sha256=qqEFbY_v1XGyFXBmwd5HSXSSHnA9wWo-Hm1vGEyIG0k,2718
 solana_agent/client/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-solana_agent/client/solana_agent.py,sha256=8mu0OdLBQnZXKS2mFrWvvY_4bykzhT73oyIxs9tmJvY,5318
+solana_agent/client/solana_agent.py,sha256=-bIVaE5p6He4d5VRzYhlAgkMzhql4EVFsujjnoweh2o,5355
 solana_agent/domains/__init__.py,sha256=HiC94wVPRy-QDJSSRywCRrhrFfTBeHjfi5z-QfZv46U,168
 solana_agent/domains/agent.py,sha256=WTo-pEc66V6D_35cpDE-kTsw1SJM-dtylPZ7em5em7Q,2659
 solana_agent/domains/routing.py,sha256=UDlgTjUoC9xIBVYu_dnf9-KG_bBgdEXAv_UtDOrYo0w,650
 solana_agent/factories/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 solana_agent/factories/agent_factory.py,sha256=mJQb1G0-gebizZvSVHm4NAxRMB1kemm2w_BAcYlN15Y,5496
 solana_agent/interfaces/__init__.py,sha256=IQs1WIM1FeKP1-kY2FEfyhol_dB-I-VAe2rD6jrVF6k,355
-solana_agent/interfaces/client/client.py,sha256=RbZWS_YaAyav56cODECMNkYL0MJwA_8bFeUreZp26qo,1697
+solana_agent/interfaces/client/client.py,sha256=btAt-bVVxsunA6rcbn0jqVmZ1JMxcF_u95CIXByO7fk,1728
 solana_agent/interfaces/plugins/plugins.py,sha256=T8HPBsekmzVwfU_Rizp-vtzAeYkMlKMYD7U9d0Wjq9c,3338
 solana_agent/interfaces/providers/data_storage.py,sha256=NqGeFvAzhz9rr-liLPRNCGjooB2EIhe-EVsMmX__b0M,1658
-solana_agent/interfaces/providers/llm.py,sha256=Fy7_iTI8sez2NVeE9lDed4W5cXx95HRy0ctEpIqr5N0,1548
+solana_agent/interfaces/providers/llm.py,sha256=_sbgSs3Sy1QAeFCB_jzw_Rjpq-N5wBY5qt6tmFYD9K4,1591
 solana_agent/interfaces/providers/memory.py,sha256=oNOH8WZXVW8assDigIWZAWiwkxbpDiKupxA2RB6tQvQ,1010
-solana_agent/interfaces/services/agent.py,sha256=7HOGcvvHTxeK-dMlqw460yqKwGd72JokhDIr3kzroVg,2120
-solana_agent/interfaces/services/query.py,sha256=PGW2w60R615og28Bw6sS1cCcBN_26KkkOsYDYclS1KQ,1347
+solana_agent/interfaces/services/agent.py,sha256=mvXl5JLiJJz0ajjVuntR-Sz8geRGs9RVqOEBsf8VzzE,2151
+solana_agent/interfaces/services/query.py,sha256=rKIYjHBeOaFFawFYduJbMRp7imYg-uRElZoizBgua00,1378
 solana_agent/interfaces/services/routing.py,sha256=UzJC-z-Q9puTWPFGEo2_CAhIxuxP5IRnze7S66NSrsI,397
 solana_agent/plugins/__init__.py,sha256=coZdgJKq1ExOaj6qB810i3rEhbjdVlrkN76ozt_Ojgo,193
 solana_agent/plugins/manager.py,sha256=Il49hXeqvu0b02pURNNp7mY8kp9_sqpi_vJIWBW5Hc0,5044
@@ -26,10 +26,10 @@ solana_agent/plugins/tools/auto_tool.py,sha256=DgES_cZ6xKSf_HJpFINpvJxrjVlk5oeqa
 solana_agent/repositories/__init__.py,sha256=fP83w83CGzXLnSdq-C5wbw9EhWTYtqE2lQTgp46-X_4,163
 solana_agent/repositories/memory.py,sha256=mrpmNSQ0D_eLebNY-cBqtecVVpIGXE7s9jCzOWEAuR4,6984
 solana_agent/services/__init__.py,sha256=ab_NXJmwYUCmCrCzuTlZ47bJZINW0Y0F5jfQ9OovidU,163
-solana_agent/services/agent.py,sha256=uLZvMl8U40H-Lbxsf6PFR4SSfggjByBUoumvB1Afduo,18441
-solana_agent/services/query.py,sha256=gUIMJaTcGUjn7TuwJHE-CHMjQIdcYUNoxqJ3duE-QUg,11278
+solana_agent/services/agent.py,sha256=ODHI66S2Y65lHrE6Kbe0zNORiz16oaMNGT3x6NoKDKo,19909
+solana_agent/services/query.py,sha256=os_LRkDIwXQuWW_zJMtm__n0Lvi-AvItdanpCs1bXv0,11362
 solana_agent/services/routing.py,sha256=PMCSG5m3uLMaHMj3dxNvNfcFZaeaDi7kMr7AEBCzwDE,6499
-solana_agent-22.0.1.dist-info/LICENSE,sha256=BnSRc-NSFuyF2s496l_4EyrwAP6YimvxWcjPiJ0J7g4,1057
-solana_agent-22.0.1.dist-info/METADATA,sha256=iUmnx3K1S69RYYoBNXE8hYBfXugafCX3r_ZMKOtJEjY,14503
-solana_agent-22.0.1.dist-info/WHEEL,sha256=fGIA9gx4Qxk2KDKeNJCbOEwSrmLtjWCwzBz351GyrPQ,88
-solana_agent-22.0.1.dist-info/RECORD,,
+solana_agent-22.0.3.dist-info/LICENSE,sha256=BnSRc-NSFuyF2s496l_4EyrwAP6YimvxWcjPiJ0J7g4,1057
+solana_agent-22.0.3.dist-info/METADATA,sha256=njD8XhVplqd_P6MtEvMHk88WvXQf7rs2nGuMyq0p4-M,14882
+solana_agent-22.0.3.dist-info/WHEEL,sha256=fGIA9gx4Qxk2KDKeNJCbOEwSrmLtjWCwzBz351GyrPQ,88
+solana_agent-22.0.3.dist-info/RECORD,,