quantalogic 0.2.7__tar.gz → 0.2.10__tar.gz

{quantalogic-0.2.7 → quantalogic-0.2.10}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: quantalogic
-Version: 0.2.7
+Version: 0.2.10
 Summary: QuantaLogic ReAct Agents
 Author: Raphaël MANSUY
 Author-email: raphael.mansuy@gmail.com
@@ -117,15 +117,25 @@ pipx install quantalogic
 The QuantaLogic CLI provides powerful command-line capabilities:
 
 ```bash
-quantalogic [OPTIONS] COMMAND [ARGS]...
-```
+Usage: quantalogic [OPTIONS] COMMAND [ARGS]...
+
+  QuantaLogic AI Assistant - A powerful AI tool for various tasks.
 
-### Options
---model-name TEXT       Model to use (default: deepseek/deepseek-chat)
---verbose               Enable verbose output
---mode TEXT             Agent mode (code/basic/interpreter/full/code-basic)
---file PATH             Path to task file
---version               Show version and exit
+Options:
+  --version                       Show version information.
+  --model-name TEXT               Specify the text model to use (litellm format,
+                                  e.g. "openrouter/deepseek-chat").
+  --vision-model-name TEXT        Specify the vision model to use (litellm format,
+                                  e.g. "openrouter/A/gpt-4o-mini").
+  --log [info|debug|warning]      Set logging level (info/debug/warning).
+  --verbose                       Enable verbose output.
+  --mode [code|basic|interpreter|full|code-basic]
+                                  Agent mode (code/search/full).
+  --help                          Show this message and exit.
+
+Commands:
+  task  Execute a task with the QuantaLogic AI Assistant.
+```
 
 ### Commands
 task    Execute a task with the QuantaLogic AI Assistant
@@ -140,33 +150,25 @@ task    Execute a task with the QuantaLogic AI Assistant
 - code-basic: Coding agent with basic reasoning
 
 #### Task Execution
+
 Tasks can be provided:
-1. Directly via --task parameter
+
+1. Directly via `task` parameter
 2. Through a file using --file parameter
 3. Interactively via standard input
 
-#### Error Handling
-The CLI provides comprehensive error handling:
-- File not found errors
-- Permission errors
-- Model configuration errors
-- Task validation errors
 
 #### Examples
 
-Basic task execution:
-```bash
-quantalogic --task "Write a Python function to calculate Fibonacci numbers"
-```
 
 Using a task file:
 ```bash
-quantalogic --file tasks/example.md --verbose
+quantalogic task --file tasks/example.md --verbose
 ```
 
 Selecting agent mode:
 ```bash
-quantalogic --mode interpreter --task "Explain quantum computing"
+quantalogic --mode interpreter task "Explain quantum computing"
 ```
 
 Interactive mode:
@@ -174,22 +176,7 @@ Interactive mode:
 quantalogic
 ```
 
-### Examples
-
-Basic task execution:
-```bash
-quantalogic --task "Write a Python function to calculate Fibonacci numbers"
-```
-
-Using a task file:
-```bash
-quantalogic --file tasks/example.md --verbose
-```
-
-Selecting agent mode:
-```bash
-quantalogic --mode interpreter --task "Explain quantum computing"
-```
+### Using QuantaLogic With code
 
 ```python
 from quantalogic import Agent
@@ -223,32 +210,6 @@ result = agent.solve_task("Create a Python function that calculates the Fibonacc
 print(result)
 ```
 
-### Command Line Interface
-
-```bash
-# Simple task execution
-quantalogic --task "Write a FastAPI endpoint that handles file uploads"
-```
-
-
-`instruction.md`
-
-```md
-Write a poem
-Translate the poem in English
-Choose 2 French Authors
-And then rewrite the translation with the style of each author
-```
-
-`
-```bash
-# Advanced usage with options
-quantalogic \
-    --model openrouter/deepseek-chat \
-    --execute-file instruction.md \
-    --verbose
-```
-
 ## 📖 Examples
 
 Here are some practical examples to help you get started:
@@ -426,7 +387,7 @@ By integrating these tools into its architecture, QuantaLogic allows agents to p
 | Script Execution       | Python Tool, Node.js Tool, Elixir Tool                                                            |
 | File Operations        | Read File Tool, Write File Tool, Edit Whole Content Tool, Replace In File Tool                   |
 | Code Analysis          | Search Definition Names Tool, Ripgrep Tool                                                        |
-| Content Generation      | LLM Tool                                                                                          |
+| Content Generation      | LLM Tool, LLMVisionTool                                                                           |
 | Utility and Management  | Download HTTP File Tool, List Directory Tool, Markitdown Tool, Unified Diff Tool                 |
 
 ---
@@ -722,7 +683,30 @@ print("Ripgrep Results:", output)
 
 ---
 
-### 14. LLM Tool
+#### 14. LLMVisionTool
+
+The **LLMVisionTool** enables processing of visual inputs using vision-language models.
+
+##### Parameters
+
+| Parameter      | Type    | Description                                                            | Example                                    |
+|----------------|---------|------------------------------------------------------------------------|--------------------------------------------|
+| `image_path`   | string  | Path to the image file to process                                      | `./path/to/image.png`                      |
+| `prompt`       | string  | The question or instruction for the vision model                       | `Describe the contents of this image`      |
+| `temperature`  | float   | Sampling temperature between 0.0 and 1.0                              | `0.7`                                      |
+
+##### Example Usage
+```python
+vision_tool = LLMVisionTool()
+response = vision_tool.execute(
+    image_path="./path/to/image.png",
+    prompt="Describe the contents of this image",
+    temperature=0.7
+)
+print("Vision Model Response:", response)
+```
+
+#### 15. LLM Tool
 
 The **LLM Tool** generates answers using a specified language model.
 
@@ -747,7 +731,7 @@ print("LLM Response:", response)
 
 ---
 
-### 15. Download HTTP File Tool
+### 16. Download HTTP File Tool
 
 The **Download HTTP File Tool** downloads a file from a specified HTTP URL.
 
@@ -767,7 +751,7 @@ print(result)
 
 ---
 
-### 16. List Directory Tool
+### 17. List Directory Tool
 
 The **List Directory Tool** lists files in a specified directory.
 
@@ -786,7 +770,7 @@ print("Directory Files:", result)
 
 ---
 
-### 17. Markitdown Tool
+### 18. Markitdown Tool
 
 The **Markitdown Tool** processes markdown files, possibly for conversion or rendering.
 
@@ -803,26 +787,6 @@ result = markitdown_tool.execute(markdown_path="./path/to/file.md")
 print("Processed Markdown Output:", result)
 ```
 
----
-
-### 18. Unified Diff Tool
-
-The **Unified Diff Tool** generates a unified diff between two texts or files.
-
-#### Parameters
-
-| Parameter    | Type   | Description                                    | Example                |
-|--------------|--------|------------------------------------------------|------------------------|
-| `original`   | string | The original content or file path.            | `old_text.txt`         |
-| `updated`    | string | The updated content or file path.             | `new_text.txt`         |
-
-#### Example Usage
-```python
-diff_tool = UnifiedDiffTool()
-result = diff_tool.execute(original="old_text.txt", updated="new_text.txt")
-print("Unified Diff Output:", result)
-```
-
 
 #### Creating Custom Tools
 
@@ -1048,3 +1012,8 @@ Copyright 2024 QuantaLogic Contributors
 
 Licensed under the Apache License, Version 2.0. See [LICENSE](LICENSE) for details.
 
+## Project Growth
+[![Star History Chart](https://api.star-history.com/svg?repos=quantalogic/quantalogic&type=Date)](https://star-history.com/#quantalogic/quantalogic&Date)
+
+Initiated with ❤️ by Raphaël MANSUY. Founder of [Quantalogic](https://www.quantalogic.app). 
+

{quantalogic-0.2.7 → quantalogic-0.2.10}/README.md

@@ -80,15 +80,25 @@ pipx install quantalogic
 The QuantaLogic CLI provides powerful command-line capabilities:
 
 ```bash
-quantalogic [OPTIONS] COMMAND [ARGS]...
-```
+Usage: quantalogic [OPTIONS] COMMAND [ARGS]...
+
+  QuantaLogic AI Assistant - A powerful AI tool for various tasks.
 
-### Options
---model-name TEXT       Model to use (default: deepseek/deepseek-chat)
---verbose               Enable verbose output
---mode TEXT             Agent mode (code/basic/interpreter/full/code-basic)
---file PATH             Path to task file
---version               Show version and exit
+Options:
+  --version                       Show version information.
+  --model-name TEXT               Specify the text model to use (litellm format,
+                                  e.g. "openrouter/deepseek-chat").
+  --vision-model-name TEXT        Specify the vision model to use (litellm format,
+                                  e.g. "openrouter/A/gpt-4o-mini").
+  --log [info|debug|warning]      Set logging level (info/debug/warning).
+  --verbose                       Enable verbose output.
+  --mode [code|basic|interpreter|full|code-basic]
+                                  Agent mode (code/search/full).
+  --help                          Show this message and exit.
+
+Commands:
+  task  Execute a task with the QuantaLogic AI Assistant.
+```
 
 ### Commands
 task    Execute a task with the QuantaLogic AI Assistant
@@ -103,33 +113,25 @@ task    Execute a task with the QuantaLogic AI Assistant
 - code-basic: Coding agent with basic reasoning
 
 #### Task Execution
+
 Tasks can be provided:
-1. Directly via --task parameter
+
+1. Directly via `task` parameter
 2. Through a file using --file parameter
 3. Interactively via standard input
 
-#### Error Handling
-The CLI provides comprehensive error handling:
-- File not found errors
-- Permission errors
-- Model configuration errors
-- Task validation errors
 
 #### Examples
 
-Basic task execution:
-```bash
-quantalogic --task "Write a Python function to calculate Fibonacci numbers"
-```
 
 Using a task file:
 ```bash
-quantalogic --file tasks/example.md --verbose
+quantalogic task --file tasks/example.md --verbose
 ```
 
 Selecting agent mode:
 ```bash
-quantalogic --mode interpreter --task "Explain quantum computing"
+quantalogic --mode interpreter task "Explain quantum computing"
 ```
 
 Interactive mode:
@@ -137,22 +139,7 @@ Interactive mode:
 quantalogic
 ```
 
-### Examples
-
-Basic task execution:
-```bash
-quantalogic --task "Write a Python function to calculate Fibonacci numbers"
-```
-
-Using a task file:
-```bash
-quantalogic --file tasks/example.md --verbose
-```
-
-Selecting agent mode:
-```bash
-quantalogic --mode interpreter --task "Explain quantum computing"
-```
+### Using QuantaLogic With code
 
 ```python
 from quantalogic import Agent
@@ -186,32 +173,6 @@ result = agent.solve_task("Create a Python function that calculates the Fibonacc
 print(result)
 ```
 
-### Command Line Interface
-
-```bash
-# Simple task execution
-quantalogic --task "Write a FastAPI endpoint that handles file uploads"
-```
-
-
-`instruction.md`
-
-```md
-Write a poem
-Translate the poem in English
-Choose 2 French Authors
-And then rewrite the translation with the style of each author
-```
-
-`
-```bash
-# Advanced usage with options
-quantalogic \
-    --model openrouter/deepseek-chat \
-    --execute-file instruction.md \
-    --verbose
-```
-
 ## 📖 Examples
 
 Here are some practical examples to help you get started:
@@ -389,7 +350,7 @@ By integrating these tools into its architecture, QuantaLogic allows agents to p
 | Script Execution       | Python Tool, Node.js Tool, Elixir Tool                                                            |
 | File Operations        | Read File Tool, Write File Tool, Edit Whole Content Tool, Replace In File Tool                   |
 | Code Analysis          | Search Definition Names Tool, Ripgrep Tool                                                        |
-| Content Generation      | LLM Tool                                                                                          |
+| Content Generation      | LLM Tool, LLMVisionTool                                                                           |
 | Utility and Management  | Download HTTP File Tool, List Directory Tool, Markitdown Tool, Unified Diff Tool                 |
 
 ---
@@ -685,7 +646,30 @@ print("Ripgrep Results:", output)
 
 ---
 
-### 14. LLM Tool
+#### 14. LLMVisionTool
+
+The **LLMVisionTool** enables processing of visual inputs using vision-language models.
+
+##### Parameters
+
+| Parameter      | Type    | Description                                                            | Example                                    |
+|----------------|---------|------------------------------------------------------------------------|--------------------------------------------|
+| `image_path`   | string  | Path to the image file to process                                      | `./path/to/image.png`                      |
+| `prompt`       | string  | The question or instruction for the vision model                       | `Describe the contents of this image`      |
+| `temperature`  | float   | Sampling temperature between 0.0 and 1.0                              | `0.7`                                      |
+
+##### Example Usage
+```python
+vision_tool = LLMVisionTool()
+response = vision_tool.execute(
+    image_path="./path/to/image.png",
+    prompt="Describe the contents of this image",
+    temperature=0.7
+)
+print("Vision Model Response:", response)
+```
+
+#### 15. LLM Tool
 
 The **LLM Tool** generates answers using a specified language model.
 
@@ -710,7 +694,7 @@ print("LLM Response:", response)
 
 ---
 
-### 15. Download HTTP File Tool
+### 16. Download HTTP File Tool
 
 The **Download HTTP File Tool** downloads a file from a specified HTTP URL.
 
@@ -730,7 +714,7 @@ print(result)
 
 ---
 
-### 16. List Directory Tool
+### 17. List Directory Tool
 
 The **List Directory Tool** lists files in a specified directory.
 
@@ -749,7 +733,7 @@ print("Directory Files:", result)
 
 ---
 
-### 17. Markitdown Tool
+### 18. Markitdown Tool
 
 The **Markitdown Tool** processes markdown files, possibly for conversion or rendering.
 
@@ -766,26 +750,6 @@ result = markitdown_tool.execute(markdown_path="./path/to/file.md")
 print("Processed Markdown Output:", result)
 ```
 
----
-
-### 18. Unified Diff Tool
-
-The **Unified Diff Tool** generates a unified diff between two texts or files.
-
-#### Parameters
-
-| Parameter    | Type   | Description                                    | Example                |
-|--------------|--------|------------------------------------------------|------------------------|
-| `original`   | string | The original content or file path.            | `old_text.txt`         |
-| `updated`    | string | The updated content or file path.             | `new_text.txt`         |
-
-#### Example Usage
-```python
-diff_tool = UnifiedDiffTool()
-result = diff_tool.execute(original="old_text.txt", updated="new_text.txt")
-print("Unified Diff Output:", result)
-```
-
 
 #### Creating Custom Tools
 
@@ -1010,3 +974,8 @@ See [CONTRIBUTING.md](CONTRIBUTING.md) for detailed guidelines.
 Copyright 2024 QuantaLogic Contributors
 
 Licensed under the Apache License, Version 2.0. See [LICENSE](LICENSE) for details.
+
+## Project Growth
+[![Star History Chart](https://api.star-history.com/svg?repos=quantalogic/quantalogic&type=Date)](https://star-history.com/#quantalogic/quantalogic&Date)
+
+Initiated with ❤️ by Raphaël MANSUY. Founder of [Quantalogic](https://www.quantalogic.app). 

{quantalogic-0.2.7 → quantalogic-0.2.10}/pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "quantalogic"
-version = "0.2.7"
+version = "0.2.10"
 description = "QuantaLogic ReAct Agents"
 authors = ["Raphaël MANSUY <raphael.mansuy@gmail.com>"]
 readme = "README.md"

{quantalogic-0.2.7 → quantalogic-0.2.10}/quantalogic/agent.py

@@ -64,6 +64,7 @@ class Agent(BaseModel):
     event_emitter: EventEmitter = EventEmitter()
     config: AgentConfig
     task_to_solve: str
+    task_to_solve_summary: str = ""
     ask_for_user_validation: Callable[[str], bool] = console_ask_for_user_validation
     last_tool_call: dict[str, Any] = {}  # Stores the last tool call information
     total_tokens: int = 0  # Total tokens in the conversation
@@ -137,6 +138,9 @@ class Agent(BaseModel):
         logger.debug(f"Solving task... {task}")
         self._reset_session(task_to_solve=task, max_iterations=max_iterations)
 
+        # Generate task summary
+        self.task_to_solve_summary = self._generate_task_summary(task)
+
         # Add system prompt to memory
         self.memory.add(Message(role="system", content=self.config.system_prompt))
 
@@ -166,7 +170,7 @@ class Agent(BaseModel):
                 self._update_total_tokens(message_history=self.memory.memory, prompt=current_prompt)
 
                 # Emit event: Task Think Start after updating total tokens
-                self._emit_event("task_think_start")
+                self._emit_event("task_think_start", {"prompt": current_prompt})
 
                 self._compact_memory_if_needed(current_prompt)
 
@@ -443,6 +447,10 @@ class Agent(BaseModel):
             "You must analyze this answer and evaluate what to do next to solve the task.\n"
             "If the step failed, take a step back and rethink your approach.\n"
             "\n"
+            "--- Task to solve summary ---\n"
+            "\n"
+            f"{self.task_to_solve_summary}"
+            "\n"
             "--- Format ---\n"
             "\n"
             "You MUST respond with exactly two XML blocks formatted in markdown:\n"
@@ -546,6 +554,7 @@ class Agent(BaseModel):
             "\n### Tools:\n"
             "-----------------------------\n"
             f"{self._get_tools_names_prompt()}\n"
+            "\n"
             "### Variables:\n"
             "-----------------------------\n"
             f"{self._get_variable_prompt()}\n"
@@ -574,7 +583,7 @@ class Agent(BaseModel):
             "\n"
             "Available variables:\n"
             "\n"
-            f"{', '.join(self.variable_store.keys())}\n"
+            f"{', '.join(self.variable_store.keys())}\n" if len(self.variable_store.keys()) > 0 else "None\n"
         )
         return prompt_use_variables
 
@@ -619,6 +628,28 @@ class Agent(BaseModel):
         self.memory.memory = memory_copy
         return summary.response
 
+    def _generate_task_summary(self, content: str) -> str:
+        """Generate a concise summary of the given content using the generative model.
+        
+        Args:
+            content (str): The content to summarize
+            
+        Returns:
+            str: Generated summary
+        """
+        try:
+            prompt = (
+                "Rewrite this task in a precise, dense, and concise manner:\n"
+                f"{content}\n"
+                "Summary should be 2-3 sentences maximum. No extra comments should be added.\n"
+            )
+            result = self.model.generate(prompt=prompt)
+            logger.debug(f"Generated summary: {result.response}")
+            return result.response
+        except Exception as e:
+            logger.error(f"Error generating summary: {str(e)}")
+            return f"Summary generation failed: {str(e)}"
+
     def _update_session_memory(self, user_content: str, assistant_content: str) -> None:
         """
         Log session messages to memory and emit events.
@@ -639,3 +670,5 @@ class Agent(BaseModel):
             "session_add_message",
             {"role": "assistant", "content": assistant_content},
         )
+        
+