quantalogic 0.31.0__py3-none-any.whl → 0.32.0__py3-none-any.whl

quantalogic/__init__.py

@@ -11,12 +11,12 @@ warnings.filterwarnings(
 )
 
 # Import public API
-from .llm import generate_completion, generate_image, count_tokens  # noqa: E402
 from .agent import Agent  # noqa: E402
-from .event_emitter import EventEmitter  # noqa: E402
-from .memory import AgentMemory, VariableMemory  # noqa: E402
 from .console_print_events import console_print_events  # noqa: E402
 from .console_print_token import console_print_token  # noqa: E402
+from .event_emitter import EventEmitter  # noqa: E402
+from .llm import count_tokens, generate_completion, generate_image  # noqa: E402
+from .memory import AgentMemory, VariableMemory  # noqa: E402
 
 __all__ = [
     "Agent",

quantalogic/agent.py

@@ -412,6 +412,9 @@ class Agent(BaseModel):
 
     def _parse_tool_usage(self, content: str) -> dict:
         """Extract tool usage from the response content."""
+        if not content or not isinstance(content, str):
+            return {}
+        
         xml_parser = ToleranceXMLParser()
         tool_names = self.tools.tool_names()
         return xml_parser.extract_elements(text=content, element_names=tool_names)
@@ -502,6 +505,7 @@ class Agent(BaseModel):
 
         # Format the response message
         formatted_response = (
+            f"Your next step: you Must now plan the next tool call to complete the based on this new observation\n"
             f"\n--- Observations for iteration {iteration} / max {self.max_iterations} ---\n"
             f"\n--- Tool execution result in ${variable_name}$ ---\n"
             f"<{variable_name}>\n{response_display}\n</{variable_name}>\n\n"
@@ -747,6 +751,8 @@ class Agent(BaseModel):
             str: Generated task summary
         """
         try:
+            if len(content) < 200:
+                return content
             prompt = (
                 "Create an ultra-concise task summary that captures ONLY: \n"
                 "1. Primary objective/purpose\n"

quantalogic/get_model_info.py

@@ -1,42 +1,83 @@
-model_info = {
-    "dashscope/qwen-max": {"max_output_tokens": 8 * 1024, "max_input_tokens": 32 * 1024},
-    "dashscope/qwen-plus": {"max_output_tokens": 8 * 1024, "max_input_tokens": 131072},
-    "dashscope/qwen-turbo": {"max_output_tokens": 8 * 1024, "max_input_tokens": 1000000},
-    "deepseek-reasoner": {"max_output_tokens": 8 * 1024, "max_input_tokens": 1024 * 128},
-    "openrouter/deepseek/deepseek-r1": {"max_output_tokens": 8 * 1024, "max_input_tokens": 1024 * 128},
-    "openrouter/mistralai/mistral-large-2411": {"max_output_tokens": 128 * 1024, "max_input_tokens": 1024 * 128},
-    "mistralai/mistral-large-2411": {"max_output_tokens": 128 * 1024, "max_input_tokens": 1024 * 128},
-}
+import loguru
+
+from quantalogic.model_info_list import model_info
+from quantalogic.model_info_litellm import litellm_get_model_max_input_tokens, litellm_get_model_max_output_tokens
+from quantalogic.utils.lm_studio_model_info import ModelInfo, get_model_list
+
+DEFAULT_MAX_OUTPUT_TOKENS = 4 * 1024  # Reasonable default for most models
+DEFAULT_MAX_INPUT_TOKENS = 32 * 1024  # Reasonable default for most models
+
+
+def validate_model_name(model_name: str) -> None:
+    if not isinstance(model_name, str) or not model_name.strip():
+        raise ValueError(f"Invalid model name: {model_name}")
 
 
 def print_model_info():
-    for model, info in model_info.items():
-        print(f"\n{model}:")
-        print(f"  Max Input Tokens: {info['max_input_tokens']:,}")
-        print(f"  Max Output Tokens: {info['max_output_tokens']:,}")
+    for info in model_info.values():
+        print(f"\n{info.model_name}:")
+        print(f"  Max Input Tokens: {info.max_input_tokens:,}")
+        print(f"  Max Output Tokens: {info.max_output_tokens:,}")
 
 
-if __name__ == "__main__":
-    print_model_info()
+def get_max_output_tokens(model_name: str) -> int:
+    """Get max output tokens with safe fallback"""
+    validate_model_name(model_name)
+
+    if model_name.startswith('lm_studio/'):
+        try:
+            models = get_model_list()
+            for model in models.data:
+                if model.id == model_name[len('lm_studio/'):]:
+                    return model.max_context_length
+        except Exception:
+            loguru.logger.warning(f"Could not fetch LM Studio model info for {model_name}, using default")
 
+    if model_name in model_info:
+        return model_info[model_name].max_output_tokens
 
-def get_max_output_tokens(model_name: str) -> int | None:
-    """Get the maximum output tokens for a given model name."""
-    return model_info.get(model_name, {}).get("max_output_tokens", None)
+    try:
+        return litellm_get_model_max_output_tokens(model_name)
+    except Exception as e:
+        loguru.logger.warning(f"Model {model_name} not found in LiteLLM registry, using default")
+        return DEFAULT_MAX_OUTPUT_TOKENS
 
 
-def get_max_input_tokens(model_name: str) -> int | None:
-    """Get the maximum input tokens for a given model name."""
-    return model_info.get(model_name, {}).get("max_input_tokens", None)
+def get_max_input_tokens(model_name: str) -> int:
+    """Get max input tokens with safe fallback"""
+    validate_model_name(model_name)
 
+    if model_name.startswith('lm_studio/'):
+        try:
+            models = get_model_list()
+            for model in models.data:
+                if model.id == model_name[len('lm_studio/'):]:
+                    return model.max_context_length
+        except Exception:
+            loguru.logger.warning(f"Could not fetch LM Studio model info for {model_name}, using default")
 
-def get_max_tokens(model_name: str) -> int | None:
-    """Get the maximum total tokens (input + output) for a given model name."""
-    model_data = model_info.get(model_name, {})
-    max_input = model_data.get("max_input_tokens")
-    max_output = model_data.get("max_output_tokens")
+    if model_name in model_info:
+        return model_info[model_name].max_input_tokens
 
-    if max_input is None or max_output is None:
-        return None
+    try:
+        return litellm_get_model_max_input_tokens(model_name)
+    except Exception:
+        loguru.logger.warning(f"Model {model_name} not found in LiteLLM registry, using default")
+        return DEFAULT_MAX_INPUT_TOKENS
 
-    return max_input + max_output
+
+def get_max_tokens(model_name: str) -> int:
+    """Get total maximum tokens (input + output)"""
+    validate_model_name(model_name)
+
+    # Get input and output tokens separately
+    input_tokens = get_max_input_tokens(model_name)
+    output_tokens = get_max_output_tokens(model_name)
+
+    return input_tokens + output_tokens
+
+
+if __name__ == "__main__":
+    print_model_info()
+    print(get_max_input_tokens("gpt-4o-mini"))
+    print(get_max_output_tokens("openrouter/openai/gpt-4o-mini"))

quantalogic/llm.py

@@ -30,18 +30,50 @@ def get_model_info(model_name: str) -> dict | None:
     return model_info.get(model_name, None)
 
 
+class ModelProviderConfig:
+    def __init__(self, prefix: str, provider: str, base_url: str, env_var: str):
+        self.prefix = prefix
+        self.provider = provider
+        self.base_url = base_url
+        self.env_var = env_var
+
+    def configure(self, model: str, kwargs: Dict[str, Any]) -> None:
+        kwargs["model"] = model.replace(self.prefix, "")
+        kwargs["custom_llm_provider"] = self.provider
+        kwargs["base_url"] = self.base_url
+        api_key = os.getenv(self.env_var)
+        if not api_key:
+            raise ValueError(f"{self.env_var} is not set in the environment variables.")
+        kwargs["api_key"] = api_key
+
+
+# Default provider configurations
+PROVIDERS = {
+    "dashscope": ModelProviderConfig(
+        prefix="dashscope/",
+        provider="openai",
+        base_url="https://dashscope-intl.aliyuncs.com/compatible-mode/v1",
+        env_var="DASHSCOPE_API_KEY"
+    ),
+    "nvidia": ModelProviderConfig(
+        prefix="nvidia/",
+        provider="openai",
+        base_url="https://integrate.api.nvidia.com/v1",
+        env_var="NVIDIA_API_KEY"
+    )
+}
+
+
 def generate_completion(**kwargs: Dict[str, Any]) -> Any:
     """Wraps litellm completion with proper type hints."""
     model = kwargs.get("model", "")
-    if model.startswith("dashscope/"):
-        # Remove prefix and configure for OpenAI-compatible endpoint
-        kwargs["model"] = model.replace("dashscope/", "")
-        kwargs["custom_llm_provider"] = "openai"  # Explicitly specify OpenAI provider
-        kwargs["base_url"] = "https://dashscope-intl.aliyuncs.com/compatible-mode/v1"
-        api_key = os.getenv("DASHSCOPE_API_KEY")
-        if not api_key:
-            raise ValueError("DASHSCOPE_API_KEY is not set in the environment variables.")
-        kwargs["api_key"] = api_key
+    
+    # Find matching provider
+    for provider_name, provider_config in PROVIDERS.items():
+        if model.startswith(provider_config.prefix):
+            provider_config.configure(model, kwargs)
+            break
+    
     return completion(**kwargs)
 
 

quantalogic/model_info.py

@@ -0,0 +1,12 @@
+from pydantic import BaseModel
+
+
+class ModelInfo(BaseModel):
+    model_name: str
+    max_input_tokens: int
+    max_output_tokens: int
+    max_cot_tokens: int | None = None
+
+
+class ModelNotFoundError(Exception):
+    """Raised when a model is not found in local registry"""

quantalogic/model_info_list.py

@@ -0,0 +1,57 @@
+from quantalogic.model_info import ModelInfo
+
+model_info = {
+    "dashscope/qwen-max": ModelInfo(
+        model_name="dashscope/qwen-max",
+        max_output_tokens=8 * 1024,
+        max_input_tokens=32 * 1024,
+    ),
+    "dashscope/qwen-plus": ModelInfo(
+        model_name="dashscope/qwen-plus",
+        max_output_tokens=8 * 1024,
+        max_input_tokens=131072,
+    ),
+    "dashscope/qwen-turbo": ModelInfo(
+        model_name="dashscope/qwen-turbo",
+        max_output_tokens=8 * 1024,
+        max_input_tokens=1000000,
+    ),
+    "deepseek-reasoner": ModelInfo(
+        model_name="deepseek-reasoner",
+        max_output_tokens=8 * 1024,
+        max_input_tokens=1024 * 128,
+    ),
+    "openrouter/deepseek/deepseek-r1": ModelInfo(
+        model_name="openrouter/deepseek/deepseek-r1",
+        max_output_tokens=8 * 1024,
+        max_input_tokens=1024 * 128,
+    ),
+    "openrouter/mistralai/mistral-large-2411": ModelInfo(
+        model_name="openrouter/mistralai/mistral-large-2411",
+        max_output_tokens=128 * 1024,
+        max_input_tokens=1024 * 128,
+    ),
+    "mistralai/mistral-large-2411": ModelInfo(
+        model_name="mistralai/mistral-large-2411",
+        max_output_tokens=128 * 1024,
+        max_input_tokens=1024 * 128,
+    ),
+    "deepseek/deepseek-chat": ModelInfo(
+        model_name="deepseek/deepseek-chat",
+        max_output_tokens=8 * 1024,
+        max_input_tokens=1024 * 64,
+    ),
+    "deepseek/deepseek-reasoner": ModelInfo(
+        model_name="deepseek/deepseek-reasoner",
+        max_output_tokens=8 * 1024,
+        max_input_tokens=1024 * 64,
+        max_cot_tokens=1024 * 32,
+    ),
+    "nvidia/deepseek-ai/deepseek-r1": ModelInfo(
+        model_name="nvidia/deepseek-ai/deepseek-r1",
+        max_output_tokens=8 * 1024,
+        max_input_tokens=1024 * 64,
+    ),
+    
+
+}

quantalogic/model_info_litellm.py

@@ -0,0 +1,70 @@
+import functools
+
+import litellm
+
+
+@functools.lru_cache(maxsize=32)
+def litellm_get_model_info(model_name: str) -> dict | None:
+    """Get model information with prefix fallback logic using only litellm.
+    
+    Args:
+        model_name: The model identifier to get information for
+        
+    Returns:
+        Dictionary containing model information
+        
+    Raises:
+        ValueError: If model info cannot be found after prefix fallbacks
+    """
+    tried_models = [model_name]
+    
+    while True:
+        try:
+            # Attempt to get model info through litellm
+            info = litellm.get_model_info(model_name)
+            if info:
+                return info
+        except Exception:
+            pass
+        
+        # Try removing one prefix level
+        parts = model_name.split("/")
+        if len(parts) <= 1:
+            break
+            
+        model_name = "/".join(parts[1:])
+        tried_models.append(model_name)
+
+    return None    
+
+
+def litellm_get_model_max_input_tokens(model_name: str) -> int | None:
+    """Get maximum input tokens for a model using litellm.
+    
+    Args:
+        model_name: The model identifier
+        
+    Returns:
+        Maximum input tokens or None if not found
+    """
+    try:
+        info = litellm_get_model_info(model_name)
+        return info.get("max_input_tokens", 8192)
+    except Exception as e:
+        return 8192  # Default for many modern models
+
+
+def litellm_get_model_max_output_tokens(model_name: str) -> int | None:
+    """Get maximum output tokens for a model using litellm.
+    
+    Args:
+        model_name: The model identifier
+        
+    Returns:
+        Maximum output tokens or None if not found
+    """
+    try:
+        info = litellm_get_model_info(model_name)
+        return info.get("max_output_tokens", 4096)
+    except Exception as e:
+        return 4096  # Conservative default

quantalogic/task_runner.py

@@ -233,6 +233,12 @@ def task_runner(
                 "memory_summary",
             ]
 
+            #def ask_continue(event: str, data: any) -> None:
+            #    ## Ask for ctrl+return
+            #    if event == "task_think_end":
+            #        ## Wait return on the keyboard
+            #        input("Press [Enter] to continue...")
+
             # Add spinner control to event handlers
             def handle_task_think_start(*args, **kwargs):
                 start_spinner(console)
@@ -246,6 +252,11 @@ def task_runner(
                 if data is not None:
                     console.print(data, end="", markup=False)
 
+            #agent.event_emitter.on(
+            #    event="task_think_end",
+            #    listener=ask_continue,
+            #)
+
             agent.event_emitter.on(
                 event=events,
                 listener=console_print_events,

quantalogic/tools/execute_bash_command_tool.py

@@ -11,7 +11,7 @@ from loguru import logger
 
 # Platform-specific imports
 try:
-    if sys.platform != 'win32':
+    if sys.platform != "win32":
         import pty
 except ImportError as e:
     logger.warning(f"Could not import platform-specific module: {e}")
@@ -40,6 +40,7 @@ class ExecuteBashCommandTool(Tool):
             description="The working directory where the command will be executed. Defaults to the current directory.",
             required=False,
             example="/path/to/directory",
+            default=os.getcwd(),
         ),
         ToolArgument(
             name="timeout",
@@ -47,6 +48,7 @@ class ExecuteBashCommandTool(Tool):
             description="Maximum time in seconds to wait for the command to complete. Defaults to 60 seconds.",
             required=False,
             example="60",
+            default="60",
         ),
     ]
 
@@ -69,16 +71,16 @@ class ExecuteBashCommandTool(Tool):
                 cwd=cwd,
                 env=env_vars,
                 text=True,
-                encoding='utf-8'
+                encoding="utf-8",
             )
 
             try:
                 stdout, stderr = process.communicate(timeout=timeout_seconds)
                 return_code = process.returncode
-                
+
                 if return_code != 0 and stderr:
                     logger.warning(f"Command failed with error: {stderr}")
-                
+
                 formatted_result = (
                     "<command_output>"
                     f" <stdout>{stdout.strip()}</stdout>"
@@ -86,11 +88,11 @@ class ExecuteBashCommandTool(Tool):
                     f"</command_output>"
                 )
                 return formatted_result
-            
+
             except subprocess.TimeoutExpired:
                 process.kill()
                 return f"Command timed out after {timeout_seconds} seconds."
-                
+
         except Exception as e:
             return f"Unexpected error executing command: {str(e)}"
 
@@ -127,7 +129,7 @@ class ExecuteBashCommandTool(Tool):
                         if proc.poll() is not None:
                             break  # Process completed but select timed out
                         raise subprocess.TimeoutExpired(command, timeout_seconds)
-                    
+
                     for fd in rlist:
                         if fd == master:
                             data = os.read(master, 1024).decode()
@@ -140,7 +142,7 @@ class ExecuteBashCommandTool(Tool):
                         elif fd == sys.stdin:
                             user_input = os.read(sys.stdin.fileno(), 1024)
                             os.write(master, user_input)
-                    
+
                     if break_loop or proc.poll() is not None:
                         while True:
                             data = os.read(master, 1024).decode()
@@ -160,7 +162,7 @@ class ExecuteBashCommandTool(Tool):
                 os.close(master)
                 proc.wait()
 
-            stdout_content = ''.join(stdout_buffer)
+            stdout_content = "".join(stdout_buffer)
             return_code = proc.returncode
             formatted_result = (
                 "<command_output>"
@@ -186,7 +188,7 @@ class ExecuteBashCommandTool(Tool):
         if env:
             env_vars.update(env)
 
-        if sys.platform == 'win32':
+        if sys.platform == "win32":
             return self._execute_windows(command, cwd, timeout_seconds, env_vars)
         else:
             if not pty:
@@ -197,4 +199,4 @@ class ExecuteBashCommandTool(Tool):
 
 if __name__ == "__main__":
     tool = ExecuteBashCommandTool()
-    print(tool.to_markdown())
+    print(tool.to_markdown())

quantalogic/utils/__init__.py

@@ -5,6 +5,7 @@ from .git_ls import git_ls
 from .get_environment import get_environment
 from .get_coding_environment import get_coding_environment
 from .get_quantalogic_rules_content import get_quantalogic_rules_file_content
+from .lm_studio_model_info import get_model_list
 
 __all__ = [
     "download_http_file",
@@ -14,4 +15,5 @@ __all__ = [
     "get_environment",
     "get_coding_environment",
     "get_quantalogic_rules_file_content",
+    "get_model_list",
 ]

quantalogic/utils/get_all_models.py

@@ -1,8 +1,8 @@
 import litellm
 
-
 from quantalogic.get_model_info import model_info
 
+
 def get_all_models() -> list[str]:
     """
     Retrieves a unified list of all model names supported by LiteLLM and Quantalogic.

quantalogic/utils/lm_studio_model_info.py

@@ -0,0 +1,48 @@
+from enum import Enum
+from typing import List, Literal, Optional
+
+from pydantic import BaseModel, Field
+
+
+class ModelType(str, Enum):
+    LLM = "llm"
+    EMBEDDINGS = "embeddings"
+    VLM = "vlm"
+
+class CompatibilityType(str, Enum):
+    MLX = "mlx"
+    GGUF = "gguf"
+
+class ModelState(str, Enum):
+    LOADED = "loaded"
+    NOT_LOADED = "not-loaded"
+
+class ModelInfo(BaseModel):
+    id: str = Field(..., description="Unique model identifier in LM Studio's namespace")
+    object: Literal["model"] = Field("model", description="Always 'model' for model objects")
+    type: ModelType = Field(..., description="Type of AI model")
+    publisher: str = Field(..., description="Organization or user who published the model")
+    arch: str = Field(..., description="Base architecture family")
+    compatibility_type: CompatibilityType = Field(..., alias="compatibility_type")
+    quantization: Optional[str] = Field(None, description="Quantization method if applicable")
+    state: ModelState = Field(..., description="Current loading state in LM Studio")
+    max_context_length: int = Field(..., alias="max_context_length", ge=0)
+    loaded_context_length: Optional[int] = Field(
+        None, 
+        alias="loaded_context_length",
+        description="Currently allocated context length (only when loaded)",
+        ge=0
+    )
+
+class ModelListResponse(BaseModel):
+    data: List[ModelInfo] = Field(..., description="List of available models")
+    object: Literal["list"] = Field("list", description="Always 'list' for list responses")
+
+def get_model_list() -> ModelListResponse:
+    """Fetch and validate model information from LM Studio's API"""
+    import requests
+    
+    response = requests.get("http://localhost:1234/api/v0/models")
+    response.raise_for_status()
+    
+    return ModelListResponse(**response.json())

quantalogic/xml_parser.py

@@ -298,5 +298,6 @@ class ToleranceXMLParser:
 
         except Exception as e:
             error_msg = f"Error extracting XML elements: {str(e)}"
+            error_msg = error_msg + f"\n{text}\n"
             logger.error(error_msg)
             raise ValueError(error_msg)

{quantalogic-0.31.0.dist-info → quantalogic-0.32.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: quantalogic
-Version: 0.31.0
+Version: 0.32.0
 Summary: QuantaLogic ReAct Agents
 Author: Raphaël MANSUY
 Author-email: raphael.mansuy@gmail.com
@@ -116,7 +116,7 @@ We created [QuantaLogic](https://www.quantalogic.app) because we saw a significa
 
 ## Usage
 
-**Usage:** `quantalogic [OPTIONS] COMMAND [ARGS]...`  
+**Usage:** `quantalogic [OPTIONS] COMMAND i[ARGS]...`  
 **Environment Variables:** Set `OPENAI_API_KEY`, `ANTHROPIC_API_KEY`, and `DEEPSEEK_API_KEY` for API integration.
 
 **Options:**
@@ -131,6 +131,7 @@ We created [QuantaLogic](https://www.quantalogic.app) because we saw a significa
 - `--help`: Show help message
 
 **Commands:**
+
 - `task`: Execute a task with the QuantaLogic AI Assistant
   - `--file PATH`: Path to task file
   - `--model-name TEXT`: Specify model
@@ -143,6 +144,25 @@ We created [QuantaLogic](https://www.quantalogic.app) because we saw a significa
   - `--compact-every-n-iteration`: Memory optimization
   - `--no-stream`: Disable streaming
 
+
+- `list-models`: List available models with optional filtering.
+  - `--search TEXT`: Filter models by name or description.
+  - `--help`: Show help message.
+
+  Example:
+  ```bash
+  quantalogic list-models --search qwen
+  ```
+
+  Output:
+  ```
+  Model Name           Description
+  -------------------  -------------------------------------------------------
+  dashscope/qwen-max   Alibaba's Qwen-Max model optimized for maximum performance
+  dashscope/qwen-plus  Alibaba's Qwen-Plus model offering balanced performance
+  ```
+
+
 ## Release Notes
 
 See our [Release Notes](RELEASE_NOTES.MD) for detailed version history and changes.
@@ -164,12 +184,25 @@ See our [Release Notes](RELEASE_NOTES.MD) for detailed version history and chang
 | openrouter/openai/gpt-4o | OPENROUTER_API_KEY | OpenAI's GPT-4o model accessible through OpenRouter platform. |
 | openrouter/mistralai/mistral-large-2411 | OPENROUTER_API_KEY | Mistral's large model optimized for complex reasoning tasks, available through OpenRouter with enhanced multilingual capabilities. |
 | mistral/mistral-large-2407 | MISTRAL_API_KEY | Mistral's high-performance model designed for enterprise-grade applications, offering advanced reasoning and multilingual support. |
+| nvidia/deepseek-ai/deepseek-r1 | NVIDIA_API_KEY | NVIDIA's DeepSeek R1 model optimized for high-performance AI tasks and advanced reasoning capabilities. |
+| lm_studio/mistral-small-24b-instruct-2501 | LM_STUDIO_API_KEY | LM Studio's Mistral Small model optimized for local inference with advanced reasoning capabilities. |
 | dashscope/qwen-max | DASHSCOPE_API_KEY | Alibaba's Qwen-Max model optimized for maximum performance and extensive reasoning capabilities. |
 | dashscope/qwen-plus | DASHSCOPE_API_KEY | Alibaba's Qwen-Plus model offering balanced performance and cost-efficiency for a variety of tasks. |
 | dashscope/qwen-turbo | DASHSCOPE_API_KEY | Alibaba's Qwen-Turbo model designed for fast and efficient responses, ideal for high-throughput scenarios. |
 
 To configure the environment API key for Quantalogic using LiteLLM, set the required environment variable for your chosen provider and any optional variables like `OPENAI_API_BASE` or `OPENROUTER_REFERRER`. Use a `.env` file or a secrets manager to securely store these keys, and load them in your code using `python-dotenv`. For advanced configurations, refer to the [LiteLLM documentation](https://docs.litellm.ai/docs/).
 
+### LM Studio Local Setup
+
+To use LM Studio with the Mistral model locally, set the following environment variables:
+
+```bash
+export LM_STUDIO_API_BASE="http://localhost:1234/v1"
+export LM_STUDIO_API_KEY="your-api-key-here"
+```
+
+Replace `http://localhost:1234/v1` with your LM Studio server URL and `your-api-key-here` with your actual API key.
+
 
 ## 📦 Installation
 

{quantalogic-0.31.0.dist-info → quantalogic-0.32.0.dist-info}/RECORD

@@ -1,5 +1,5 @@
-quantalogic/__init__.py,sha256=s8G7kW8JI_WP201l950vIHxtLHk0V7mERsHuIR--P9w,925
-quantalogic/agent.py,sha256=2qgx5mJGx12otW2KzPoM0ZSugVxSTP3sXnvilXcvnrQ,33020
+quantalogic/__init__.py,sha256=Su8CnOEdqKu4zTytjiP9P5olg-oIDuUA3fMWM1WUdRY,925
+quantalogic/agent.py,sha256=VqxM_VizDpwH3SLAnoSFMjH4a1ijBVOHaI6rXahHP7M,33289
 quantalogic/agent_config.py,sha256=SIRVSF0kkrYfvtyHiMCJhnm_nYqJCD2p1pN-reMIy24,7868
 quantalogic/agent_factory.py,sha256=HWKwN_DN57EPmME-hoCD2uJE0DqsPCzGU_V7nq54XzI,5284
 quantalogic/coding_agent.py,sha256=Z7ik6LUvLKDnaW9Ax1iZGC7p1WMnlYEUIlE5lkBP414,4975
@@ -9,11 +9,14 @@ quantalogic/console_print_token.py,sha256=qSU-3kmoZk4T5-1ybrEBi8tIXDPcz7eyWKhGh3
 quantalogic/docs_cli.py,sha256=3giVbUpespB9ZdTSJ955A3BhcOaBl5Lwsn1AVy9XAeY,1663
 quantalogic/event_emitter.py,sha256=jqot2g4JRXc88K6PW837Oqxbf7shZfO-xdPaUWmzupk,7901
 quantalogic/generative_model.py,sha256=ut_BFy4BqDxNqUXVbM8e_C_CzwNuJkvGWRsbpbKaees,13423
-quantalogic/get_model_info.py,sha256=nAvb_x4MnZei9CvTxUwqFUtFKkTp4-qiKB7ZWZxLxp8,1762
+quantalogic/get_model_info.py,sha256=_9Nb9JQ09HZzT-_gZUSvl4Er7uCXs5ys36sIBa-8DXA,3005
 quantalogic/interactive_text_editor.py,sha256=1vW4poJl7SItRGEeGQgtCFcmRDXmfCM8PE-uBtDBJuE,16658
-quantalogic/llm.py,sha256=yhuC1b5TCXojDXZEJK_PMcWUUxzrpI-gwzlIszAdJMM,4677
+quantalogic/llm.py,sha256=Nk2Dn1lJh1-323Fs7ADfR9ov_eAoJOEEnzyGswZSbJI,5460
 quantalogic/main.py,sha256=__-4pX2pgoSFvt-aLdp6Qlrq55_SrwP_l8u2uTaQbjg,9262
 quantalogic/memory.py,sha256=zbtRuM05jaS2lJll-92dt5JfYVLERnF_m_9xqp2x-k0,6304
+quantalogic/model_info.py,sha256=j7QqvjEFQDGpDOgQs8uTkVyI3a50Oa_nrsQjyxizTLc,272
+quantalogic/model_info_list.py,sha256=bJu2ohQFgZOwmcqydfh5oi5R1G8ZY7jtQlERsR1Z47s,1816
+quantalogic/model_info_litellm.py,sha256=m1Yt4SIiOBRWLx7S8f8k4fcTiKJZKtOvcPN_QvQ_Oxk,1880
 quantalogic/model_names.py,sha256=UZlz25zG9B2dpfwdw_e1Gw5qFsKQ7iME9FJh9Ts4u6s,938
 quantalogic/prompts.py,sha256=M-7rCaQoylnwxedhvy7VmQdgBG6TT1vmcf8_UzPTyY0,4035
 quantalogic/search_agent.py,sha256=EA_FAPP0dVuUbJ_lAGKfYq1FIJ6oLYzGMgKLMvBL4ZQ,2472
@@ -26,7 +29,7 @@ quantalogic/server/static/js/event_visualizer.js,sha256=eFkkWyNZw3zOZlF18kxbfsWq
 quantalogic/server/static/js/quantalogic.js,sha256=x7TrlZGR1Y0WLK2DWl1xY847BhEWMPnL0Ua7KtOldUc,22311
 quantalogic/server/templates/index.html,sha256=nDnXJoQEm1vXbhXtgaYk0G5VXj0wwzE6KrqEDhHFpj4,7773
 quantalogic/task_file_reader.py,sha256=AMIJoeVY9Hhu0dBJ-C5EyaOFsXLkhn2oBhVs-WTnnLk,1460
-quantalogic/task_runner.py,sha256=Loa0hr-Bge-kVnTpoGuV7RotFdiAKQL_zj4OB4S6gko,9643
+quantalogic/task_runner.py,sha256=6kL7o0br2YU8FNKyq-rKJo1oGKZKl73vtATpHusPIWQ,10038
 quantalogic/tool_manager.py,sha256=Uh-ufrJPufHqDUrFwKlXw3MOsVGc_4lQxuc6cRvZ7wU,7186
 quantalogic/tools/__init__.py,sha256=pTirT5UBynuTkAzFYebu7ttGAMP3_A0idFvDp6lGZJQ,2146
 quantalogic/tools/agent_tool.py,sha256=MXCXxWHRch7VK4UWhtRP1jeI8Np9Ne2CUGo8vm1oZiM,3064
@@ -35,7 +38,7 @@ quantalogic/tools/download_http_file_tool.py,sha256=wTfanbXjIRi5-qrbluuLvNmDNhvm
 quantalogic/tools/duckduckgo_search_tool.py,sha256=xVaEb_SUK5NL3lwMQXj1rGQYYvNT-td-qaB9QCes27Q,7014
 quantalogic/tools/edit_whole_content_tool.py,sha256=nXmpAvojvqvAcqNMy1kUKZ1ocboky_ZcnCR4SNCSPgw,2360
 quantalogic/tools/elixir_tool.py,sha256=fzPPtAW-Koy9KB0r5k2zV1f1U0WphL-LXPPOBkeNkug,7652
-quantalogic/tools/execute_bash_command_tool.py,sha256=kl3RSOZCOc-U52dwd0h6BxXvjMlAX7D0Bo2-HkCOcxo,6908
+quantalogic/tools/execute_bash_command_tool.py,sha256=YK_cMuODJDOYheZKGmlpZTxJdVbimFLCUlND2_zmyMg,6869
 quantalogic/tools/generate_database_report_tool.py,sha256=QbZjtmegGEOEZAIa-CSeBo5O9dYBZTk_PWrumyFUg1Q,1890
 quantalogic/tools/grep_app_tool.py,sha256=BDxygwx7WCbqbiP2jmSRnIsoIUVYG5A4SKzId524ys4,19957
 quantalogic/tools/input_question_tool.py,sha256=UoTlNhdmdr-eyiVtVCG2qJe_R4bU_ag-DzstSdmYkvM,1848
@@ -72,24 +75,25 @@ quantalogic/tools/utils/create_sample_database.py,sha256=Aus9xRLGfQfsYnxsAkJ5CW-
 quantalogic/tools/utils/generate_database_report.py,sha256=0D-5fWOfpAh1jEcld5OTQP5x6XkJE5jpNY6FyHv1L2s,10345
 quantalogic/tools/wikipedia_search_tool.py,sha256=bdZ_0dYTxpEfU04tBFsatnLM5P9Z3kAZgKQEjsopJLA,5405
 quantalogic/tools/write_file_tool.py,sha256=_mx9_Zjg2oMAAVzlcHEKjZVZUxQVgbRfcoMKgWnoZcg,3764
-quantalogic/utils/__init__.py,sha256=Ltq7tzLuHCl9BpCvfRVA9Sjrtp1RJesrn7G980lbl_c,563
+quantalogic/utils/__init__.py,sha256=E442CJQuTohKzgI0Wrd4NZEpKascFjz6F4Vy8Y1c_0Y,634
 quantalogic/utils/ask_user_validation.py,sha256=F0jkbFJVXAImcSSP7op6dov5i80hRvZGRvBHbfcZrxg,340
 quantalogic/utils/check_version.py,sha256=grxTfJE85GMue1OAk8z8_q8tjEJxQ8RO6fN3fJ_qedg,1136
 quantalogic/utils/download_http_file.py,sha256=FTN3brq9WvCFvuBX-lYAhjsdYTzQT4m9m2vqlcyjkNk,3472
-quantalogic/utils/get_all_models.py,sha256=GGhonVHUS7MeS6eogmkEwZETSVgeGU9pWCLTzR0XxAU,544
+quantalogic/utils/get_all_models.py,sha256=Ol4e60MwZiJhu8HZ2i_RpIumLmFYYrncB1X9q1KEQh0,544
 quantalogic/utils/get_coding_environment.py,sha256=oMK5ZanOqX_SFaJxUZQGlsAAaiLUgJufCJYDrHnHPuQ,584
 quantalogic/utils/get_environment.py,sha256=7wWruSHYTUlnQWW27qU3WFYZnncqqqdofsxAsUU7lhw,875
 quantalogic/utils/get_quantalogic_rules_content.py,sha256=fnEFTyClXzpI0MLaM-gB9R6l4CJlu2NnaYiR09ciJC8,673
 quantalogic/utils/git_ls.py,sha256=_k6QIQtc0aM1bsG340jBp4VrdevbcH8Pg2CV4r9oHok,5264
+quantalogic/utils/lm_studio_model_info.py,sha256=1eDvZ-I9W8AZbCch1l5rdiSpUxL7qMnfZItdFZkmAWs,1819
 quantalogic/utils/read_file.py,sha256=tSRVHk8dIP4nNLL89v5kRki4hOTjVyjbmuEb2zwvwCY,2077
 quantalogic/utils/read_http_text_content.py,sha256=n3IayT5KcqctIVVF2gOQQAMf3Ow6eenlVgfXTpLcQbw,4410
 quantalogic/version.py,sha256=ea_cRutaQk5_lwlLbUUvPFuOT7Of7-gAsDl7wdveS-g,107
 quantalogic/version_check.py,sha256=cttR1lR3OienGLl7NrK1Te1fhDkqSjCci7HC1vFUTSY,1627
 quantalogic/welcome_message.py,sha256=IXMhem8h7srzNUwvw8G_lmEkHU8PFfote021E_BXmVk,3039
-quantalogic/xml_parser.py,sha256=uMLQNHTRCg116FwcjRoquZmSwVtE4LEH-6V2E3RD-dA,11466
+quantalogic/xml_parser.py,sha256=8yDxvKzAEnefNwUAR-wjerMDOj5T5cxak4WPIA83SBw,11516
 quantalogic/xml_tool_parser.py,sha256=Vz4LEgDbelJynD1siLOVkJ3gLlfHsUk65_gCwbYJyGc,3784
-quantalogic-0.31.0.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-quantalogic-0.31.0.dist-info/METADATA,sha256=SQnNSbR3pqL1-i7F2dUruuLT650tM-qp35rktOO0FC8,22262
-quantalogic-0.31.0.dist-info/WHEEL,sha256=Nq82e9rUAnEjt98J6MlVmMCZb-t9cYE2Ir1kpBmnWfs,88
-quantalogic-0.31.0.dist-info/entry_points.txt,sha256=h74O_Q3qBRCrDR99qvwB4BpBGzASPUIjCfxHq6Qnups,183
-quantalogic-0.31.0.dist-info/RECORD,,
+quantalogic-0.32.0.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+quantalogic-0.32.0.dist-info/METADATA,sha256=w5kJUS98ou0kVeY9fy-WImrW-W9wWZFOXmIQx4xK-_M,23461
+quantalogic-0.32.0.dist-info/WHEEL,sha256=Nq82e9rUAnEjt98J6MlVmMCZb-t9cYE2Ir1kpBmnWfs,88
+quantalogic-0.32.0.dist-info/entry_points.txt,sha256=h74O_Q3qBRCrDR99qvwB4BpBGzASPUIjCfxHq6Qnups,183
+quantalogic-0.32.0.dist-info/RECORD,,