feat: update formatting tools to match new analysis structure

- Remove handling of deprecated `significant_lines` field
- Improve rich formatting of lyrics using Panel
- Add `conclusion` style for conclusion section
- Fix error with missing `neutral` style
- Remove unused Markdown import
- Enhance lyrics formatting while preserving `quote` style

Gradio_UI.py

@@ -268,7 +268,7 @@ class GradioUI:
         import gradio as gr
 
         # Define instruction text
-        instructions = """
+        ui_tip_text = """
         # 🎵 Song Meaning Bot 🎶
         ### How to Use:
         1. Paste song title in the input field.  
@@ -277,7 +277,7 @@ class GradioUI:
         """
 
         with gr.Blocks(fill_height=True) as demo:
-            gr.Markdown(instructions)
+            gr.Markdown(ui_tip_text)
             stored_messages = gr.State([])
             file_uploads_log = gr.State([])
             chatbot = gr.Chatbot(

app.py

@@ -36,7 +36,7 @@ def main():
 
     # If using Ollama, we need to specify the API base URL
     # Initialize the LLM model based on configuration
-    model_id = get_model_id(is_test=is_test)
+    model_id = get_model_id(use_local=is_test)
     logger.info(f"Initializing with model: {model_id}")
     if is_test:
         api_base = get_ollama_api_base()

config.py

@@ -24,13 +24,28 @@ def load_api_keys():
     """Load API keys from environment variables."""
     # Gemini API is the default
     os.environ["GEMINI_API_KEY"] = os.getenv("GEMINI_API_KEY")
+    os.environ["OPENROUTER_API_KEY"] = os.getenv("OPENROUTER_API_KEY")
 
-def get_model_id(is_test=True):
-    """Get the appropriate model ID based on configuration."""
-    if is_test:
-        return "ollama/gemma3:4b"  # Using local Ollama with Gemma 3:4B instead of Claude
-    else:
+def get_model_id(use_local=True, provider="ollama"):
+    """Get the appropriate model ID based on configuration.
+    
+    Args:
+        use_local: If True, use test configuration (local development).
+               If False, use production configuration.
+        provider: Model provider ('ollama', 'gemini', 'openrouter')
+    
+    Returns:
+        String with model ID for the specified provider.
+    """
+    if provider == "ollama":
+        return "ollama/gemma3:4b"  # Using local Ollama with Gemma 3:4B
+    elif provider == "gemini":
         return "gemini/gemini-2.0-flash"
+    elif provider == "openrouter":
+        return "openrouter/google/gemini-2.0-flash-lite-preview-02-05:free"  # OpenRouter Claude 3 Opus
+    else:
+        # Default fallback
+        return "ollama/gemma3:4b" if use_local else "gemini/gemini-2.0-flash"
 
 def get_ollama_api_base():
     """Get the API base URL for Ollama."""

prompts/prompts_hf.yaml

@@ -1,4 +1,4 @@
-"system_prompt": |-
+system_prompt: |-
   You are an expert assistant who can solve any task using code blobs. You will be given a task to solve as best you can.
   To do so, you have been given access to a list of tools: these tools are basically Python functions which you can call with code.
   To solve the task, you must plan forward to proceed in a series of steps, in a cycle of 'Thought:', 'Code:', and 'Observation:' sequences.
@@ -172,8 +172,8 @@
   10. Don't give up! You're in charge of solving the task, not providing directions to solve it.
 
   Now Begin! If you solve the task correctly, you will receive a reward of $1,000,000.
-"planning":
-  "initial_facts": |-
+planning:
+  initial_facts: |-
     Below I will present you a task.
 
     You will now build a comprehensive preparatory survey of which facts we have at our disposal and which ones we still need.
@@ -196,7 +196,13 @@
     ### 2. Facts to look up
     ### 3. Facts to derive
     Do not add anything else.
-  "initial_plan": |-
+
+    Here is the task:
+    ```
+    {{task}}
+    ```
+    Now begin!
+  initial_plan : |-
     You are a world expert at making efficient plans to solve any task using a set of carefully crafted tools.
 
     Now for the given task, develop a step-by-step high-level plan taking into account the above inputs and list of facts.
@@ -234,7 +240,7 @@
     ```
 
     Now begin! Write your plan below.
-  "update_facts_pre_messages": |-
+  update_facts_pre_messages: |-
     You are a world expert at gathering known and unknown facts based on a conversation.
     Below you will find a task, and a history of attempts made to solve the task. You will have to produce a list of these:
     ### 1. Facts given in the task
@@ -242,7 +248,7 @@
     ### 3. Facts still to look up
     ### 4. Facts still to derive
     Find the task and history below:
-  "update_facts_post_messages": |-
+  update_facts_post_messages: |-
     Earlier we've built a list of facts.
     But since in your previous steps you may have learned useful new facts or invalidated some false ones.
     Please update your list of facts based on the previous history, and provide these headings:
@@ -252,7 +258,7 @@
     ### 4. Facts still to derive
 
     Now write your new list of facts below.
-  "update_plan_pre_messages": |-
+  update_plan_pre_messages: |-
     You are a world expert at making efficient plans to solve any task using a set of carefully crafted tools.
 
     You have been given a task:
@@ -263,7 +269,7 @@
     Find below the record of what has been tried so far to solve it. Then you will be asked to make an updated plan to solve the task.
     If the previous tries so far have met some success, you can make an updated plan based on these actions.
     If you are stalled, you can make a completely new plan starting from scratch.
-  "update_plan_post_messages": |-
+  update_plan_post_messages: |-
     You're still working towards solving this task:
     ```
     {{task}}
@@ -299,23 +305,29 @@
     After writing the final step of the plan, write the '\n<end_plan>' tag and stop there.
 
     Now write your new plan below.
-"managed_agent":
-  "task": |-
-    You're a helpful agent named '{{name}}'.
-    You have been submitted this task by your manager.
-    ---
-    Task:
-    {{task}}
-    ---
-    You're helping your manager solve a wider task: so make sure to not provide a one-line answer, but give as much information as possible to give them a clear understanding of the answer.
-
-    Your final_answer WILL HAVE to contain these parts:
-    ### 1. Task outcome (short version):
-    ### 2. Task outcome (extremely detailed version):
-    ### 3. Additional context (if relevant):
-
-    Put all these in your final_answer tool, everything that you do not pass as an argument to final_answer will be lost.
-    And even if your task resolution is not successful, please return as much context as possible, so that your manager can act upon this feedback.
-  "report": |-
-    Here is the final answer from your managed agent '{{name}}':
-    {{final_answer}}
+managed_agent:
+  task: |-
+      You're a helpful agent named '{{name}}'.
+      You have been submitted this task by your manager.
+      ---
+      Task:
+      {{task}}
+      ---
+      You're helping your manager solve a wider task: so make sure to not provide a one-line answer, but give as much information as possible to give them a clear understanding of the answer.
+
+      Your final_answer WILL HAVE to contain these parts:
+      ### 1. Task outcome (short version):
+      ### 2. Task outcome (extremely detailed version):
+      ### 3. Additional context (if relevant):
+
+      Put all these in your final_answer tool, everything that you do not pass as an argument to final_answer will be lost.
+      And even if your task resolution is not successful, please return as much context as possible, so that your manager can act upon this feedback.
+  report: |-
+      Here is the final answer from your managed agent '{{name}}':
+      {{final_answer}}
+final_answer:
+  pre_messages: |-
+    An agent tried to answer a user query but it got stuck and failed to do so. You are tasked with providing an answer instead. Here is the agent's memory:
+  post_messages: |-
+    Based on the above, please provide an answer to the following user request:
+    {{task}}

pyproject.toml

@@ -11,4 +11,5 @@ dependencies = [
     "loguru>=0.7.3",
     "pyyaml>=6.0.2",
     "smolagents>=1.9.2",
+    "tenacity>=9.0.0",
 ]

run_single_agent.py

@@ -0,0 +1,41 @@
+import os
+
+from smolagents import LiteLLMModel
+
+from agents.single_agent import create_single_agent
+from loguru import logger
+from config import get_model_id, get_ollama_api_base, setup_logger, load_api_keys
+
+setup_logger()
+load_api_keys()
+
+# Set environment variables for API keys if needed
+os.environ["GEMINI_API_KEY"] = str(os.getenv("GEMINI_API_KEY"))
+
+
+use_local = False
+
+# If using Ollama, we need to specify the API base URL
+# Initialize the LLM model based on configuration
+model_id = "openrouter/google/gemini-2.0-flash-lite-preview-02-05:free"
+logger.info(f"Initializing with model: {model_id}")
+if use_local:
+    api_base = get_ollama_api_base()
+    logger.info(f"Using Ollama API base: {api_base}")
+    model = LiteLLMModel(model_id=model_id, api_base=api_base)
+else:
+    model = LiteLLMModel(model_id=model_id)
+
+# If the agent does not answer, the model is overloaded, please use another model or the following Hugging Face Endpoint that also contains qwen2.5 coder:
+# model_id='https://pflgm2locj2t89co.us-east-1.aws.endpoints.huggingface.cloud' 
+
+# Prompt the user for the song name
+song_data = "RCHP - On Mercury"
+
+agent = create_single_agent(model)
+
+# Agent execution
+agent.run(f"""
+1. Find and extract the lyrics of the song, {song_data}. Don't try to scrape from azlyrics.com or genius.com, others are ok.
+2. Perform deep lyrics analysis and return full lyrics and analysis results in a pretty human-readable format.
+""")

tools/formatting_tools.py

@@ -9,7 +9,6 @@ from rich.console import Console
 from rich.panel import Panel
 from rich.text import Text
 from rich.table import Table
-from rich.markdown import Markdown
 from rich.theme import Theme
 from rich.box import ROUNDED
 from rich.console import Group
@@ -37,6 +36,22 @@ class FormatAnalysisResultsTool(Tool):
         Returns:
             A formatted string representation of the analysis
         """
+        # Expected JSON structure from analysis_tools.py:
+        # {
+        #     "summary": "Overall analysis of the song vibes, meaning and mood",
+        #     "main_themes": ["theme1", "theme2", ...],
+        #     "mood": "The overall mood/emotion of the song",
+        #     "sections_analysis": [
+        #         {
+        #             "section_type": "verse/chorus/bridge/etc.",
+        #             "section_number": 1,
+        #             "lines": ["line1", "line2", ...],
+        #             "analysis": "Analysis of this section whith respect to the overall theme"
+        #         },
+        #         ...
+        #     ],
+        #     "conclusion": "The song vibes and concepts of the underlying meaning"
+        # }
         try:
             # Parse the JSON string into a Python dictionary if it's a string
             if isinstance(analysis_json, str):
@@ -59,7 +74,8 @@ class FormatAnalysisResultsTool(Tool):
                         "negative": "red",
                         "neutral": "magenta",
                         "quote": "italic yellow",
-                        "metadata": "dim white"
+                        "metadata": "dim white",
+                        "conclusion": "bold magenta"  # Add style for conclusion
                     })
                     
                     # Apply the theme to our console
@@ -103,9 +119,19 @@ class FormatAnalysisResultsTool(Tool):
                             section_content = []
                             
                             if lines:
-                                lyrics_text = "\n".join([f"> [quote]{line}[/]" for line in lines])
+                                # Format lyrics in a more readable way
                                 section_content.append(Text("Lyrics:", style="bold blue"))
-                                section_content.append(Markdown(lyrics_text))
+                                # Форматируем каждую строку лирики с стилем quote
+                                lyrics_lines = []
+                                for line in lines:
+                                    lyrics_lines.append(f"[quote]{line}[/]")
+                                
+                                lyrics_panel = Panel(
+                                    "\n".join(lyrics_lines),
+                                    border_style="blue",
+                                    padding=(1, 2)
+                                )
+                                section_content.append(lyrics_panel)
                             
                             section_content.append(Text("Analysis:", style="bold blue"))
                             section_content.append(Text(section_analysis))
@@ -117,25 +143,12 @@ class FormatAnalysisResultsTool(Tool):
                                 border_style="cyan"
                             ))
                     
-                    # Significant lines
-                    sig_lines = analysis.get("significant_lines", [])
-                    if sig_lines:
-                        console.print("\n[heading]Significant Lines[/]")
-                        
-                        for i, line_data in enumerate(sig_lines):
-                            line = line_data.get("line", "")
-                            significance = line_data.get("significance", "")
-                            
-                            console.print(Panel(
-                                f"[quote]\"{line}\"[/]\n\n[bold blue]Significance:[/] {significance}",
-                                title=f"[highlight]Key Line #{i+1}[/]",
-                                border_style="highlight"
-                            ))
+                    # We no longer have significant_lines in the new format
                     
                     # Conclusion
                     conclusion = analysis.get("conclusion", "No conclusion available")
                     console.print("\n[heading]Conclusion[/]")
-                    console.print(Panel(conclusion, border_style="neutral"))
+                    console.print(Panel(conclusion, border_style="magenta"))
                     
                     # Export the rich text as a string
                     return console.export_text()
@@ -198,20 +211,7 @@ class FormatAnalysisResultsTool(Tool):
                         formatted_text.append(section_analysis)
                         formatted_text.append("")
                 
-                # Significant lines
-                sig_lines = analysis.get("significant_lines", [])
-                if sig_lines:
-                    formatted_text.append("SIGNIFICANT LINES")
-                    formatted_text.append("=================")
-                    
-                    for i, line_data in enumerate(sig_lines):
-                        line = line_data.get("line", "")
-                        significance = line_data.get("significance", "")
-                        
-                        formatted_text.append(f"Key Line #{i+1}:")
-                        formatted_text.append(f'"{line}"')
-                        formatted_text.append(f"Significance: {significance}")
-                        formatted_text.append("")
+                # We no longer have significant_lines in the new format
                 
                 # Conclusion
                 conclusion = analysis.get("conclusion", "No conclusion available")