solana-agent 28.1.0__tar.gz → 28.2.0__tar.gz

{solana_agent-28.1.0 → solana_agent-28.2.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: solana-agent
-Version: 28.1.0
+Version: 28.2.0
 Summary: AI Agents for Solana
 License: MIT
 Keywords: solana,solana ai,solana agent,ai,ai agent,ai agents
@@ -23,7 +23,9 @@ Requires-Dist: pinecone (>=6.0.2,<7.0.0)
 Requires-Dist: pydantic (>=2)
 Requires-Dist: pymongo (>=4.12.0,<5.0.0)
 Requires-Dist: pypdf (>=5.4.0,<6.0.0)
+Requires-Dist: rich (>=13)
 Requires-Dist: scrubadub (>=2.0.1,<3.0.0)
+Requires-Dist: typer (>=0.15.2,<0.16.0)
 Requires-Dist: zep-cloud (>=2.10.1,<3.0.0)
 Project-URL: Documentation, https://docs.solana-agent.com
 Project-URL: Homepage, https://solana-agent.com
@@ -76,7 +78,7 @@ Build your AI agents in three lines of code!
 * Fast AI responses
 * Solana Ecosystem Integration via [AgentiPy](https://github.com/niceberginc/agentipy)
 * MCP tool usage with first-class support for [Zapier](https://zapier.com/mcp)
-* Integrated observability and tracing via [Pydantic Logfire](https://logfire.pydantic.dev/)
+* Integrated observability and tracing via [Pydantic Logfire](https://pydantic.dev/logfire)
 * Designed for a multi-agent swarm 
 * Seamless text and audio streaming with real-time multi-modal processing
 * Persistent memory that preserves context across all agent interactions
@@ -101,7 +103,7 @@ Build your AI agents in three lines of code!
 * [Pinecone](https://pinecone.io) - Knowledge Base (optional)
 * [AgentiPy](https://agentipy.fun) - Solana Ecosystem (optional)
 * [Zapier](https://zapier.com) - App Integrations (optional)
-* [Pydantic Logfire](https://logfire.pydantic.dev) - Observability and Tracing (optional)
+* [Pydantic Logfire](https://pydantic.dev/logfire) - Observability and Tracing (optional)
 
 ### AI Models Used
 
@@ -284,6 +286,47 @@ async for response in solana_agent.process("user123", audio_content, audio_input
     print(response, end="")
 ```
 
+### Command Line Interface (CLI)
+
+Solana Agent includes a command-line interface (CLI) for text-based chat using a configuration file.
+
+Ensure you have a valid configuration file (e.g., `config.json`) containing at least your OpenAI API key and agent definitions.
+
+**./config.json**
+```json
+{
+    "openai": {
+        "api_key": "your-openai-api-key"
+    },
+    "agents": [
+        {
+            "name": "default_agent",
+            "instructions": "You are a helpful AI assistant.",
+            "specialization": "general"
+        }
+    ]
+}
+```
+
+Also ensure that you have `pip install uv` to call `uvx`.
+
+```bash
+uvx solana-agent [OPTIONS]
+
+Options:
+
+--user-id TEXT: The user ID for the conversation (default: cli_user).
+--config TEXT: Path to the configuration JSON file (default: config.json).
+--prompt TEXT: Optional system prompt override for the agent.
+--help: Show help message and exit.
+
+# Using default config.json and user_id
+uvx solana-agent
+
+# Specifying user ID and config path
+uvx solana-agent --user-id my_cli_session --config ./my_agent_config.json
+```
+
 ## Optional Feature Configs
 
 ### Business Alignment

{solana_agent-28.1.0 → solana_agent-28.2.0}/README.md

@@ -44,7 +44,7 @@ Build your AI agents in three lines of code!
 * Fast AI responses
 * Solana Ecosystem Integration via [AgentiPy](https://github.com/niceberginc/agentipy)
 * MCP tool usage with first-class support for [Zapier](https://zapier.com/mcp)
-* Integrated observability and tracing via [Pydantic Logfire](https://logfire.pydantic.dev/)
+* Integrated observability and tracing via [Pydantic Logfire](https://pydantic.dev/logfire)
 * Designed for a multi-agent swarm 
 * Seamless text and audio streaming with real-time multi-modal processing
 * Persistent memory that preserves context across all agent interactions
@@ -69,7 +69,7 @@ Build your AI agents in three lines of code!
 * [Pinecone](https://pinecone.io) - Knowledge Base (optional)
 * [AgentiPy](https://agentipy.fun) - Solana Ecosystem (optional)
 * [Zapier](https://zapier.com) - App Integrations (optional)
-* [Pydantic Logfire](https://logfire.pydantic.dev) - Observability and Tracing (optional)
+* [Pydantic Logfire](https://pydantic.dev/logfire) - Observability and Tracing (optional)
 
 ### AI Models Used
 
@@ -252,6 +252,47 @@ async for response in solana_agent.process("user123", audio_content, audio_input
     print(response, end="")
 ```
 
+### Command Line Interface (CLI)
+
+Solana Agent includes a command-line interface (CLI) for text-based chat using a configuration file.
+
+Ensure you have a valid configuration file (e.g., `config.json`) containing at least your OpenAI API key and agent definitions.
+
+**./config.json**
+```json
+{
+    "openai": {
+        "api_key": "your-openai-api-key"
+    },
+    "agents": [
+        {
+            "name": "default_agent",
+            "instructions": "You are a helpful AI assistant.",
+            "specialization": "general"
+        }
+    ]
+}
+```
+
+Also ensure that you have `pip install uv` to call `uvx`.
+
+```bash
+uvx solana-agent [OPTIONS]
+
+Options:
+
+--user-id TEXT: The user ID for the conversation (default: cli_user).
+--config TEXT: Path to the configuration JSON file (default: config.json).
+--prompt TEXT: Optional system prompt override for the agent.
+--help: Show help message and exit.
+
+# Using default config.json and user_id
+uvx solana-agent
+
+# Specifying user ID and config path
+uvx solana-agent --user-id my_cli_session --config ./my_agent_config.json
+```
+
 ## Optional Feature Configs
 
 ### Business Alignment

{solana_agent-28.1.0 → solana_agent-28.2.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "solana-agent"
-version = "28.1.0"
+version = "28.2.0"
 description = "AI Agents for Solana"
 authors = ["Bevan Hunt <bevan@bevanhunt.com>"]
 license = "MIT"
@@ -35,6 +35,8 @@ llama-index-embeddings-openai = "^0.3.1"
 pypdf = "^5.4.0"
 scrubadub = "^2.0.1"
 logfire = "^3.14.0"
+typer = "^0.15.2"
+rich = ">=13"
 
 [tool.poetry.group.dev.dependencies]
 pytest = "^8.3.5"
@@ -49,6 +51,9 @@ sphinx-autobuild = "^2024.10.3"
 mongomock = "^4.3.0"
 ruff = "^0.11.6"
 
+[tool.poetry.scripts]
+solana-agent = "solana_agent.cli:app"
+
 [build-system]
 requires = ["poetry-core>=1.0.0"]
 build-backend = "poetry.core.masonry.api"

{solana_agent-28.1.0 → solana_agent-28.2.0}/solana_agent/adapters/openai_adapter.py

@@ -4,6 +4,7 @@ LLM provider adapters for the Solana Agent system.
 These adapters implement the LLMProvider interface for different LLM services.
 """
 
+import logging
 from typing import AsyncGenerator, List, Literal, Optional, Type, TypeVar
 
 from openai import AsyncOpenAI
@@ -14,6 +15,9 @@ import logfire
 
 from solana_agent.interfaces.providers.llm import LLMProvider
 
+# Setup logger for this module
+logger = logging.getLogger(__name__)
+
 T = TypeVar("T", bound=BaseModel)
 
 DEFAULT_CHAT_MODEL = "gpt-4.1"
@@ -35,12 +39,10 @@ class OpenAIAdapter(LLMProvider):
             try:
                 logfire.configure(token=logfire_api_key)
                 self.logfire = True
-                print("Logfire configured successfully.")  # Optional: confirmation log
+                logger.info("Logfire configured successfully.")  # Use logger.info
             except Exception as e:
-                print(
-                    f"Failed to configure Logfire: {e}"
-                )  # Log error if configuration fails
-                self.logfire = False  # Ensure logfire is False if config fails
+                logger.error(f"Failed to configure Logfire: {e}")  # Use logger.error
+                self.logfire = False
 
         self.parse_model = DEFAULT_PARSE_MODEL
         self.text_model = DEFAULT_CHAT_MODEL
@@ -79,7 +81,8 @@ class OpenAIAdapter(LLMProvider):
             Audio bytes as they become available
         """
         try:
-            logfire.instrument_openai(self.client)
+            if self.logfire:  # Instrument only if logfire is enabled
+                logfire.instrument_openai(self.client)
             async with self.client.audio.speech.with_streaming_response.create(
                 model=self.tts_model,
                 voice=voice,
@@ -91,17 +94,8 @@ class OpenAIAdapter(LLMProvider):
                     yield chunk
 
         except Exception as e:
-            print(f"Error in text_to_speech: {str(e)}")
-            import traceback
-
-            print(traceback.format_exc())
-            yield b""  # Return empty bytes on error
-
-        except Exception as e:
-            print(f"Error in text_to_speech: {str(e)}")
-            import traceback
-
-            print(traceback.format_exc())
+            # Log the exception with traceback
+            logger.exception(f"Error in text_to_speech: {e}")
             yield b""  # Return empty bytes on error
 
     async def transcribe_audio(
@@ -121,7 +115,8 @@ class OpenAIAdapter(LLMProvider):
             Transcript text chunks as they become available
         """
         try:
-            logfire.instrument_openai(self.client)
+            if self.logfire:  # Instrument only if logfire is enabled
+                logfire.instrument_openai(self.client)
             async with self.client.audio.transcriptions.with_streaming_response.create(
                 model=self.transcription_model,
                 file=(f"file.{input_format}", audio_bytes),
@@ -132,10 +127,8 @@ class OpenAIAdapter(LLMProvider):
                     yield chunk
 
         except Exception as e:
-            print(f"Error in transcribe_audio: {str(e)}")
-            import traceback
-
-            print(traceback.format_exc())
+            # Log the exception with traceback
+            logger.exception(f"Error in transcribe_audio: {e}")
             yield f"I apologize, but I encountered an error transcribing the audio: {str(e)}"
 
     async def generate_text(
@@ -177,12 +170,16 @@ class OpenAIAdapter(LLMProvider):
                 full_text = response.choices[0].message.content
                 return full_text  # Return the complete string
             else:
-                print("Received non-streaming response with no content.")
+                logger.warning(
+                    "Received non-streaming response with no content."
+                )  # Use logger.warning
                 return ""  # Return empty string if no content
 
         except Exception as e:
-            # Log the error and return an error message string
-            print(f"Error in generate_text: {e}")
+            # Log the exception and return an error message string
+            logger.exception(f"Error in generate_text: {e}")
+            # Consider returning a more informative error string or raising
+            return f"Error generating text: {e}"
 
     async def parse_structured_output(
         self,
@@ -208,56 +205,67 @@ class OpenAIAdapter(LLMProvider):
             if self.logfire:
                 logfire.instrument_openai(client)
 
-            if model:
-                self.parse_model = model
+            # Use the provided model or the default parse model
+            current_parse_model = model or self.parse_model
 
             patched_client = instructor.from_openai(client, mode=Mode.TOOLS_STRICT)
 
             # Use instructor's structured generation with function calling
             response = await patched_client.chat.completions.create(
-                model=self.parse_model,
+                model=current_parse_model,  # Use the determined model
                 messages=messages,
                 response_model=model_class,
                 max_retries=2,  # Automatically retry on validation errors
             )
             return response
         except Exception as e:
-            print(f"Error with instructor parsing (TOOLS_STRICT mode): {e}")
+            logger.warning(
+                f"Instructor parsing (TOOLS_STRICT mode) failed: {e}"
+            )  # Log warning
 
             try:
+                # Determine client again for fallback
                 if api_key and base_url:
                     client = AsyncOpenAI(api_key=api_key, base_url=base_url)
                 else:
                     client = self.client
 
-                if model:
-                    self.parse_model = model
+                if self.logfire:  # Instrument again if needed
+                    logfire.instrument_openai(client)
+
+                # Use the provided model or the default parse model
+                current_parse_model = model or self.parse_model
 
                 # First fallback: Try regular JSON mode
+                logger.info("Falling back to instructor JSON mode.")  # Log info
                 patched_client = instructor.from_openai(client, mode=Mode.JSON)
                 response = await patched_client.chat.completions.create(
-                    model=self.parse_model,
+                    model=current_parse_model,  # Use the determined model
                     messages=messages,
                     response_model=model_class,
                     max_retries=1,
                 )
                 return response
             except Exception as json_error:
-                print(f"JSON mode fallback also failed: {json_error}")
+                logger.warning(
+                    f"Instructor JSON mode fallback also failed: {json_error}"
+                )  # Log warning
 
                 try:
+                    # Determine client again for final fallback
                     if api_key and base_url:
                         client = AsyncOpenAI(api_key=api_key, base_url=base_url)
                     else:
                         client = self.client
 
-                    if self.logfire:
+                    if self.logfire:  # Instrument again if needed
                         logfire.instrument_openai(client)
 
-                    if model:
-                        self.parse_model = model
+                    # Use the provided model or the default parse model
+                    current_parse_model = model or self.parse_model
 
                     # Final fallback: Manual extraction with a detailed prompt
+                    logger.info("Falling back to manual JSON extraction.")  # Log info
                     fallback_system_prompt = f"""
                     {system_prompt}
 
@@ -269,7 +277,7 @@ class OpenAIAdapter(LLMProvider):
 
                     # Regular completion without instructor
                     completion = await client.chat.completions.create(
-                        model=self.parse_model,
+                        model=current_parse_model,  # Use the determined model
                         messages=[
                             {"role": "system", "content": fallback_system_prompt},
                             {"role": "user", "content": prompt},
@@ -284,7 +292,10 @@ class OpenAIAdapter(LLMProvider):
                     return model_class.model_validate_json(json_str)
 
                 except Exception as fallback_error:
-                    print(f"All fallback methods failed: {fallback_error}")
+                    # Log the final exception with traceback
+                    logger.exception(
+                        f"All structured output fallback methods failed: {fallback_error}"
+                    )
                     raise ValueError(
                         f"Failed to generate structured output: {e}. All fallbacks failed."
                     ) from e
@@ -303,6 +314,8 @@ class OpenAIAdapter(LLMProvider):
             A list of floats representing the embedding vector.
         """
         if not text:
+            # Log error instead of raising immediately, let caller handle empty input if needed
+            logger.error("Attempted to embed empty text.")
             raise ValueError("Text cannot be empty")
 
         try:
@@ -313,7 +326,7 @@ class OpenAIAdapter(LLMProvider):
             # Replace newlines with spaces as recommended by OpenAI
             text = text.replace("\n", " ")
 
-            if self.logfire:
+            if self.logfire:  # Instrument only if logfire is enabled
                 logfire.instrument_openai(self.client)
 
             response = await self.client.embeddings.create(
@@ -323,11 +336,13 @@ class OpenAIAdapter(LLMProvider):
             if response.data and response.data[0].embedding:
                 return response.data[0].embedding
             else:
+                # Log warning about unexpected response structure
+                logger.warning(
+                    "Failed to retrieve embedding from OpenAI response structure."
+                )
                 raise ValueError("Failed to retrieve embedding from OpenAI response")
 
         except Exception as e:
-            print(f"Error generating embedding: {e}")
-            import traceback
-
-            print(traceback.format_exc())
-            raise
+            # Log the exception with traceback before raising
+            logger.exception(f"Error generating embedding: {e}")
+            raise  # Re-raise the original exception