solana-agent 27.5.0__tar.gz → 28.1.0__tar.gz

{solana_agent-27.5.0 → solana_agent-28.1.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: solana-agent
-Version: 27.5.0
+Version: 28.1.0
 Summary: AI Agents for Solana
 License: MIT
 Keywords: solana,solana ai,solana agent,ai,ai agent,ai agents
@@ -17,6 +17,7 @@ Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
 Requires-Dist: instructor (>=1.7.9,<2.0.0)
 Requires-Dist: llama-index-core (>=0.12.30,<0.13.0)
 Requires-Dist: llama-index-embeddings-openai (>=0.3.1,<0.4.0)
+Requires-Dist: logfire (>=3.14.0,<4.0.0)
 Requires-Dist: openai (>=1.75.0,<2.0.0)
 Requires-Dist: pinecone (>=6.0.2,<7.0.0)
 Requires-Dist: pydantic (>=2)
@@ -59,9 +60,11 @@ Build your AI agents in three lines of code!
 * Intelligent Routing
 * Business Alignment
 * Extensible Tooling
+* Automatic Tool Workflows
 * Knowledge Base
 * MCP Support
 * Guardrails
+* Pydantic Logfire
 * Tested & Secure
 * Built in Python
 * Powers [CometHeart](https://cometheart.com)
@@ -73,6 +76,7 @@ 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/)
 * 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
@@ -84,26 +88,28 @@ Build your AI agents in three lines of code!
 * Assigned tools are utilized by agents automatically and effectively
 * Integrated Knowledge Base with semantic search and automatic PDF chunking
 * Input and output guardrails for content filtering, safety, and data sanitization
+* Automatic sequential tool workflows allowing agents to chain multiple tools
 
 ## Stack
 
 ### Tech
 
 * [Python](https://python.org) - Programming Language
-* [OpenAI](https://openai.com), [Google](https://ai.google.dev), [xAI](https://x.ai) - LLM Providers
+* [OpenAI](https://openai.com) - AI Model Provider
 * [MongoDB](https://mongodb.com) - Conversational History (optional)
 * [Zep Cloud](https://getzep.com) - Conversational Memory (optional)
 * [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)
 
-### LLMs
+### AI Models Used
 
-* [gpt-4.1-mini](https://platform.openai.com/docs/models/gpt-4.1-mini) (agent)
+* [gpt-4.1](https://platform.openai.com/docs/models/gpt-4.1) (agent)
 * [gpt-4.1-nano](https://platform.openai.com/docs/models/gpt-4.1-nano) (router)
 * [text-embedding-3-large](https://platform.openai.com/docs/models/text-embedding-3-large) or [text-embedding-3-small](https://platform.openai.com/docs/models/text-embedding-3-small) (embedding)
 * [tts-1](https://platform.openai.com/docs/models/tts-1) (audio TTS)
 * [gpt-4o-mini-transcribe](https://platform.openai.com/docs/models/gpt-4o-mini-transcribe) (audio transcription)
-* [gemini-2.5-flash-preview](https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview) (optional)
-* [grok-3-mini-fast-beta](https://docs.x.ai/docs/models#models-and-pricing) (optional)
 
 ## Installation
 
@@ -113,13 +119,13 @@ You can install Solana Agent using pip:
 
 ## Flows
 
-In both flows of single and multiple agents - it is one user query to one agent using one tool (if needed).
+In both flows of single and multiple agents - it is one user query to one agent using one or many tools (if needed).
 
-An agent can have multiple tools and will choose the best one to answer the user query.
+An agent can have multiple tools and will choose the best ones to fulfill the user's query.
 
-Routing is determined by optimal domain expertise of the agent for the user query.
+Routing is determined by optimal domain expertise of the agent for the user's query.
 
-When the agent uses a tool it feeds the tool output back to itself to generate the final response.
+When the agent uses tools it feeds the tools output back to itself to generate the final response.
 
 This is important as tools generally output unstructured and unformatted data that the agent needs to prepare for the user.
 
@@ -128,13 +134,13 @@ Keep this in mind while designing your agentic systems using Solana Agent.
 ```ascii
                        Single Agent                                     
                                                                         
-     ┌────────┐        ┌─────────┐        ┌────────┐                    
-     │        │        │         │        │        │                    
-     │        │        │         │        │        │                    
-     │  User  │◄──────►│  Agent  │◄──────►│  Tool  │                    
-     │        │        │         │        │        │                    
-     │        │        │         │        │        │                    
-     └────────┘        └─────────┘        └────────┘                    
+     ┌────────┐        ┌─────────┐        ┌────────-┐                    
+     │        │        │         │        │         │                    
+     │        │        │         │        │         │                    
+     │  User  │◄──────►│  Agent  │◄──────►│  Tools  │                    
+     │        │        │         │        │         │                    
+     │        │        │         │        │         │                    
+     └────────┘        └─────────┘        └────────-┘                    
                                                                         
                                                                         
                                                                         
@@ -142,13 +148,13 @@ Keep this in mind while designing your agentic systems using Solana Agent.
                                                                         
                       Multiple Agents                                   
                                                                         
-     ┌────────┐        ┌──────────┐        ┌─────────┐        ┌────────┐
-     │        │        │          │        │         │        │        │
-     │        │        │          │        │         │        │        │
-┌───►│  User  ├───────►│  Router  ├───────►│  Agent  │◄──────►│  Tool  │
-│    │        │        │          │        │         │        │        │
-│    │        │        │          │        │         │        │        │
-│    └────────┘        └──────────┘        └────┬────┘        └────────┘
+     ┌────────┐        ┌──────────┐        ┌─────────┐        ┌────────-┐
+     │        │        │          │        │         │        │         │
+     │        │        │          │        │         │        │         │
+┌───►│  User  ├───────►│  Router  ├───────►│  Agent  │◄──────►│  Tools  │
+│    │        │        │          │        │         │        │         │
+│    │        │        │          │        │         │        │         │
+│    └────────┘        └──────────┘        └────┬────┘        └────────-┘
 │                                               │                       
 │                                               │                       
 │                                               │                       
@@ -319,26 +325,12 @@ config = {
 }
 ```
 
-### Gemini
-
-This allows Gemini to replace OpenAI for agent and router.
-
-```python
-config = {
-    "gemini": {
-        "api_key": "your-gemini-api-key",
-    },
-}
-```
-
-### Grok
-
-This allows Grok to replace OpenAI (or Gemini) for agent.
+### Observability and Tracing
 
 ```python
 config = {
-    "grok": {
-        "api_key": "your-grok-api-key",
+    "logfire": {
+        "api_key": "your-logfire-write-token",
     },
 }
 ```
@@ -533,8 +525,8 @@ class MyOutputGuardrail(OutputGuardrail):
 
 Tools can be used from plugins like Solana Agent Kit (sakit) or via inline tools. Tools available via plugins integrate automatically with Solana Agent.
 
-* Agents can only call one tool per response
-* Agents choose the best tool for the job
+* Agents can use multiple tools per response and should apply the right sequential order (like send an email to bob@bob.com with the latest news on Solana)
+* Agents choose the best tools for the job
 * Solana Agent doesn't use OpenAI function calling (tools) as they don't support async functions
 * Solana Agent tools are async functions
 

{solana_agent-27.5.0 → solana_agent-28.1.0}/README.md

@@ -28,9 +28,11 @@ Build your AI agents in three lines of code!
 * Intelligent Routing
 * Business Alignment
 * Extensible Tooling
+* Automatic Tool Workflows
 * Knowledge Base
 * MCP Support
 * Guardrails
+* Pydantic Logfire
 * Tested & Secure
 * Built in Python
 * Powers [CometHeart](https://cometheart.com)
@@ -42,6 +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/)
 * 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
@@ -53,26 +56,28 @@ Build your AI agents in three lines of code!
 * Assigned tools are utilized by agents automatically and effectively
 * Integrated Knowledge Base with semantic search and automatic PDF chunking
 * Input and output guardrails for content filtering, safety, and data sanitization
+* Automatic sequential tool workflows allowing agents to chain multiple tools
 
 ## Stack
 
 ### Tech
 
 * [Python](https://python.org) - Programming Language
-* [OpenAI](https://openai.com), [Google](https://ai.google.dev), [xAI](https://x.ai) - LLM Providers
+* [OpenAI](https://openai.com) - AI Model Provider
 * [MongoDB](https://mongodb.com) - Conversational History (optional)
 * [Zep Cloud](https://getzep.com) - Conversational Memory (optional)
 * [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)
 
-### LLMs
+### AI Models Used
 
-* [gpt-4.1-mini](https://platform.openai.com/docs/models/gpt-4.1-mini) (agent)
+* [gpt-4.1](https://platform.openai.com/docs/models/gpt-4.1) (agent)
 * [gpt-4.1-nano](https://platform.openai.com/docs/models/gpt-4.1-nano) (router)
 * [text-embedding-3-large](https://platform.openai.com/docs/models/text-embedding-3-large) or [text-embedding-3-small](https://platform.openai.com/docs/models/text-embedding-3-small) (embedding)
 * [tts-1](https://platform.openai.com/docs/models/tts-1) (audio TTS)
 * [gpt-4o-mini-transcribe](https://platform.openai.com/docs/models/gpt-4o-mini-transcribe) (audio transcription)
-* [gemini-2.5-flash-preview](https://ai.google.dev/gemini-api/docs/models#gemini-2.5-flash-preview) (optional)
-* [grok-3-mini-fast-beta](https://docs.x.ai/docs/models#models-and-pricing) (optional)
 
 ## Installation
 
@@ -82,13 +87,13 @@ You can install Solana Agent using pip:
 
 ## Flows
 
-In both flows of single and multiple agents - it is one user query to one agent using one tool (if needed).
+In both flows of single and multiple agents - it is one user query to one agent using one or many tools (if needed).
 
-An agent can have multiple tools and will choose the best one to answer the user query.
+An agent can have multiple tools and will choose the best ones to fulfill the user's query.
 
-Routing is determined by optimal domain expertise of the agent for the user query.
+Routing is determined by optimal domain expertise of the agent for the user's query.
 
-When the agent uses a tool it feeds the tool output back to itself to generate the final response.
+When the agent uses tools it feeds the tools output back to itself to generate the final response.
 
 This is important as tools generally output unstructured and unformatted data that the agent needs to prepare for the user.
 
@@ -97,13 +102,13 @@ Keep this in mind while designing your agentic systems using Solana Agent.
 ```ascii
                        Single Agent                                     
                                                                         
-     ┌────────┐        ┌─────────┐        ┌────────┐                    
-     │        │        │         │        │        │                    
-     │        │        │         │        │        │                    
-     │  User  │◄──────►│  Agent  │◄──────►│  Tool  │                    
-     │        │        │         │        │        │                    
-     │        │        │         │        │        │                    
-     └────────┘        └─────────┘        └────────┘                    
+     ┌────────┐        ┌─────────┐        ┌────────-┐                    
+     │        │        │         │        │         │                    
+     │        │        │         │        │         │                    
+     │  User  │◄──────►│  Agent  │◄──────►│  Tools  │                    
+     │        │        │         │        │         │                    
+     │        │        │         │        │         │                    
+     └────────┘        └─────────┘        └────────-┘                    
                                                                         
                                                                         
                                                                         
@@ -111,13 +116,13 @@ Keep this in mind while designing your agentic systems using Solana Agent.
                                                                         
                       Multiple Agents                                   
                                                                         
-     ┌────────┐        ┌──────────┐        ┌─────────┐        ┌────────┐
-     │        │        │          │        │         │        │        │
-     │        │        │          │        │         │        │        │
-┌───►│  User  ├───────►│  Router  ├───────►│  Agent  │◄──────►│  Tool  │
-│    │        │        │          │        │         │        │        │
-│    │        │        │          │        │         │        │        │
-│    └────────┘        └──────────┘        └────┬────┘        └────────┘
+     ┌────────┐        ┌──────────┐        ┌─────────┐        ┌────────-┐
+     │        │        │          │        │         │        │         │
+     │        │        │          │        │         │        │         │
+┌───►│  User  ├───────►│  Router  ├───────►│  Agent  │◄──────►│  Tools  │
+│    │        │        │          │        │         │        │         │
+│    │        │        │          │        │         │        │         │
+│    └────────┘        └──────────┘        └────┬────┘        └────────-┘
 │                                               │                       
 │                                               │                       
 │                                               │                       
@@ -288,26 +293,12 @@ config = {
 }
 ```
 
-### Gemini
-
-This allows Gemini to replace OpenAI for agent and router.
-
-```python
-config = {
-    "gemini": {
-        "api_key": "your-gemini-api-key",
-    },
-}
-```
-
-### Grok
-
-This allows Grok to replace OpenAI (or Gemini) for agent.
+### Observability and Tracing
 
 ```python
 config = {
-    "grok": {
-        "api_key": "your-grok-api-key",
+    "logfire": {
+        "api_key": "your-logfire-write-token",
     },
 }
 ```
@@ -502,8 +493,8 @@ class MyOutputGuardrail(OutputGuardrail):
 
 Tools can be used from plugins like Solana Agent Kit (sakit) or via inline tools. Tools available via plugins integrate automatically with Solana Agent.
 
-* Agents can only call one tool per response
-* Agents choose the best tool for the job
+* Agents can use multiple tools per response and should apply the right sequential order (like send an email to bob@bob.com with the latest news on Solana)
+* Agents choose the best tools for the job
 * Solana Agent doesn't use OpenAI function calling (tools) as they don't support async functions
 * Solana Agent tools are async functions
 

{solana_agent-27.5.0 → solana_agent-28.1.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "solana-agent"
-version = "27.5.0"
+version = "28.1.0"
 description = "AI Agents for Solana"
 authors = ["Bevan Hunt <bevan@bevanhunt.com>"]
 license = "MIT"
@@ -34,6 +34,7 @@ llama-index-core = "^0.12.30"
 llama-index-embeddings-openai = "^0.3.1"
 pypdf = "^5.4.0"
 scrubadub = "^2.0.1"
+logfire = "^3.14.0"
 
 [tool.poetry.group.dev.dependencies]
 pytest = "^8.3.5"

{solana_agent-27.5.0 → solana_agent-28.1.0}/solana_agent/adapters/openai_adapter.py

@@ -10,12 +10,13 @@ from openai import AsyncOpenAI
 from pydantic import BaseModel
 import instructor
 from instructor import Mode
+import logfire
 
 from solana_agent.interfaces.providers.llm import LLMProvider
 
 T = TypeVar("T", bound=BaseModel)
 
-DEFAULT_CHAT_MODEL = "gpt-4.1-mini"
+DEFAULT_CHAT_MODEL = "gpt-4.1"
 DEFAULT_PARSE_MODEL = "gpt-4.1-nano"
 DEFAULT_EMBEDDING_MODEL = "text-embedding-3-large"
 DEFAULT_EMBEDDING_DIMENSIONS = 3072
@@ -26,8 +27,21 @@ DEFAULT_TTS_MODEL = "tts-1"
 class OpenAIAdapter(LLMProvider):
     """OpenAI implementation of LLMProvider with web search capabilities."""
 
-    def __init__(self, api_key: str):
+    def __init__(self, api_key: str, logfire_api_key: Optional[str] = None):
         self.client = AsyncOpenAI(api_key=api_key)
+
+        self.logfire = False
+        if logfire_api_key:
+            try:
+                logfire.configure(token=logfire_api_key)
+                self.logfire = True
+                print("Logfire configured successfully.")  # Optional: confirmation log
+            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
+
         self.parse_model = DEFAULT_PARSE_MODEL
         self.text_model = DEFAULT_CHAT_MODEL
         self.transcription_model = DEFAULT_TRANSCRIPTION_MODEL
@@ -65,6 +79,7 @@ class OpenAIAdapter(LLMProvider):
             Audio bytes as they become available
         """
         try:
+            logfire.instrument_openai(self.client)
             async with self.client.audio.speech.with_streaming_response.create(
                 model=self.tts_model,
                 voice=voice,
@@ -106,6 +121,7 @@ class OpenAIAdapter(LLMProvider):
             Transcript text chunks as they become available
         """
         try:
+            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),
@@ -129,45 +145,44 @@ class OpenAIAdapter(LLMProvider):
         api_key: Optional[str] = None,
         base_url: Optional[str] = None,
         model: Optional[str] = None,
-    ) -> AsyncGenerator[str, None]:  # pragma: no cover
-        """Generate text from OpenAI models."""
+    ) -> str:  # pragma: no cover
+        """Generate text from OpenAI models as a single string."""
         messages = []
-
         if system_prompt:
             messages.append({"role": "system", "content": system_prompt})
-
         messages.append({"role": "user", "content": prompt})
 
-        # Prepare request parameters
+        # Prepare request parameters - stream is always False now
         request_params = {
             "messages": messages,
-            "stream": True,
-            "model": self.text_model,
+            "stream": False,  # Hardcoded to False
+            "model": model or self.text_model,
         }
 
+        # Determine client based on provided api_key/base_url
         if api_key and base_url:
             client = AsyncOpenAI(api_key=api_key, base_url=base_url)
         else:
             client = self.client
 
-        if model:
-            request_params["model"] = model
+        if self.logfire:
+            logfire.instrument_openai(client)
 
         try:
+            # Make the non-streaming API call
             response = await client.chat.completions.create(**request_params)
 
-            async for chunk in response:
-                if chunk.choices:
-                    if chunk.choices[0].delta.content:
-                        text = chunk.choices[0].delta.content
-                        yield text
+            # Handle non-streaming response
+            if response.choices and response.choices[0].message.content:
+                full_text = response.choices[0].message.content
+                return full_text  # Return the complete string
+            else:
+                print("Received non-streaming response with no content.")
+                return ""  # Return empty string if no content
 
         except Exception as e:
-            print(f"Error in generate_text: {str(e)}")
-            import traceback
-
-            print(traceback.format_exc())
-            yield f"I apologize, but I encountered an error: {str(e)}"
+            # Log the error and return an error message string
+            print(f"Error in generate_text: {e}")
 
     async def parse_structured_output(
         self,
@@ -190,6 +205,9 @@ class OpenAIAdapter(LLMProvider):
             else:
                 client = self.client
 
+            if self.logfire:
+                logfire.instrument_openai(client)
+
             if model:
                 self.parse_model = model
 
@@ -233,6 +251,9 @@ class OpenAIAdapter(LLMProvider):
                     else:
                         client = self.client
 
+                    if self.logfire:
+                        logfire.instrument_openai(client)
+
                     if model:
                         self.parse_model = model
 
@@ -292,6 +313,9 @@ class OpenAIAdapter(LLMProvider):
             # Replace newlines with spaces as recommended by OpenAI
             text = text.replace("\n", " ")
 
+            if self.logfire:
+                logfire.instrument_openai(self.client)
+
             response = await self.client.embeddings.create(
                 input=[text], model=embedding_model, dimensions=embedding_dimensions
             )

{solana_agent-27.5.0 → solana_agent-28.1.0}/solana_agent/factories/agent_factory.py

@@ -78,6 +78,10 @@ class SolanaAgentFactory:
         # Create adapters
 
         if "mongo" in config:
+            if "connection_string" not in config["mongo"]:
+                raise ValueError("MongoDB connection string is required.")
+            if "database" not in config["mongo"]:
+                raise ValueError("MongoDB database name is required.")
             db_adapter = MongoDBAdapter(
                 connection_string=config["mongo"]["connection_string"],
                 database_name=config["mongo"]["database"],
@@ -85,9 +89,21 @@ class SolanaAgentFactory:
         else:
             db_adapter = None
 
-        llm_adapter = OpenAIAdapter(
-            api_key=config["openai"]["api_key"],
-        )
+        if "logfire" in config:
+            if "api_key" not in config["logfire"]:
+                raise ValueError("Pydantic Logfire API key is required.")
+            if "openai" not in config or "api_key" not in config["openai"]:
+                raise ValueError("OpenAI API key is required.")
+            llm_adapter = OpenAIAdapter(
+                api_key=config["openai"]["api_key"],
+                logfire_api_key=config["logfire"].get("api_key"),
+            )
+        else:
+            if "openai" not in config or "api_key" not in config["openai"]:
+                raise ValueError("OpenAI API key is required.")
+            llm_adapter = OpenAIAdapter(
+                api_key=config["openai"].get("api_key"),
+            )
 
         # Create business mission if specified in config
         business_mission = None
@@ -130,90 +146,19 @@ class SolanaAgentFactory:
             f"Loaded {len(input_guardrails)} input guardrails and {len(output_guardrails)} output guardrails."
         )
 
-        if (
-            "gemini" in config
-            and "api_key" in config["gemini"]
-            and "grok" not in config
-        ):
-            # Create primary services
-            agent_service = AgentService(
-                llm_provider=llm_adapter,
-                business_mission=business_mission,
-                config=config,
-                api_key=config["gemini"]["api_key"],
-                base_url="https://generativelanguage.googleapis.com/v1beta/openai/",
-                model="gemini-2.5-flash-preview-04-17",
-                output_guardrails=output_guardrails,
-            )
-
-            # Create routing service
-            routing_service = RoutingService(
-                llm_provider=llm_adapter,
-                agent_service=agent_service,
-                api_key=config["gemini"]["api_key"],
-                base_url="https://generativelanguage.googleapis.com/v1beta/openai/",
-                model="gemini-2.5-flash-preview-04-17",
-            )
-
-        elif (
-            "gemini" in config
-            and "api_key" in config["gemini"]
-            and "grok" in config
-            and "api_key" in config["grok"]
-        ):
-            # Create primary services
-            agent_service = AgentService(
-                llm_provider=llm_adapter,
-                business_mission=business_mission,
-                config=config,
-                api_key=config["grok"]["api_key"],
-                base_url="https://api.x.ai/v1",
-                model="grok-3-mini-fast-beta",
-                output_guardrails=output_guardrails,
-            )
-            # Create routing service
-            routing_service = RoutingService(
-                llm_provider=llm_adapter,
-                agent_service=agent_service,
-                api_key=config["gemini"]["api_key"],
-                base_url="https://generativelanguage.googleapis.com/v1beta/openai/",
-                model="gemini-2.5-flash-preview-04-17",
-            )
-
-        elif (
-            "grok" in config and "api_key" in config["grok"] and "gemini" not in config
-        ):
-            # Create primary services
-            agent_service = AgentService(
-                llm_provider=llm_adapter,
-                business_mission=business_mission,
-                config=config,
-                api_key=config["grok"]["api_key"],
-                base_url="https://api.x.ai/v1",
-                model="grok-3-mini-fast-beta",
-                output_guardrails=output_guardrails,
-            )
-
-            # Create routing service
-            routing_service = RoutingService(
-                llm_provider=llm_adapter,
-                agent_service=agent_service,
-            )
-
-        else:
-            # Create primary services
-            agent_service = AgentService(
-                llm_provider=llm_adapter,
-                business_mission=business_mission,
-                config=config,
-                output_guardrails=output_guardrails,
-            )
+        # Create primary services
+        agent_service = AgentService(
+            llm_provider=llm_adapter,
+            business_mission=business_mission,
+            config=config,
+            output_guardrails=output_guardrails,
+        )
 
-            # Create routing service
-            routing_service = RoutingService(
-                llm_provider=llm_adapter,
-                agent_service=agent_service,
-            )
+        # Create routing service
+        routing_service = RoutingService(
+            llm_provider=llm_adapter,
+            agent_service=agent_service,
+        )
 
         # Debug the agent service tool registry
         print(

{solana_agent-27.5.0 → solana_agent-28.1.0}/solana_agent/interfaces/providers/llm.py

@@ -25,7 +25,7 @@ class LLMProvider(ABC):
         api_key: Optional[str] = None,
         base_url: Optional[str] = None,
         model: Optional[str] = None,
-    ) -> AsyncGenerator[str, None]:
+    ) -> str:
         """Generate text from the language model."""
         pass