updated final assignment

.env

@@ -0,0 +1 @@
+SPACE_ID = "abhi1294/Final_Assignment_Template"

agent.py

@@ -0,0 +1,78 @@
+from __future__ import annotations
+
+from dataclasses import dataclass
+from typing import Optional
+
+from prompts import build_solver_prompt
+from tools import TaskFileTool
+from utils import extract_final_answer, normalize_final_answer
+
+
+@dataclass
+class AgentConfig:
+    api_base_url: str = "https://agents-course-unit4-scoring.hf.space"
+    max_context_chars: int = 12000
+
+
+class SubmissionAgent:
+    """
+    V1 agent for the Hugging Face Agents Course Unit 4 final project.
+
+    Goals:
+    - Accept a benchmark question and optional task_id
+    - Load attached task-file context when available
+    - Return ONLY the final answer string
+    - Stay framework-agnostic for now so we can plug in any LLM later
+    """
+
+    def __init__(self, llm_client=None, config: Optional[AgentConfig] = None):
+        self.llm_client = llm_client
+        self.config = config or AgentConfig()
+        self.task_file_tool = TaskFileTool(api_base_url=self.config.api_base_url)
+
+    def __call__(self, question: str, task_id: Optional[str] = None) -> str:
+        """
+        Main entry point used by app.py.
+        """
+        context = self._load_context(task_id=task_id)
+        raw_output = self._solve(question=question, context=context)
+        final_answer = extract_final_answer(raw_output)
+        return normalize_final_answer(final_answer)
+
+    def _load_context(self, task_id: Optional[str]) -> str:
+        """
+        Try to fetch and read any task-linked file.
+        Safe fallback: empty context.
+        """
+        if not task_id:
+            return ""
+
+        try:
+            file_text = self.task_file_tool.get_task_context(task_id=task_id)
+            if not file_text:
+                return ""
+
+            return file_text[: self.config.max_context_chars]
+        except Exception:
+            return ""
+
+    def _solve(self, question: str, context: str) -> str:
+        """
+        Solve the question with either:
+        1) a plugged-in LLM client, or
+        2) a safe fallback so the app does not crash during setup.
+
+        The LLM client is expected to expose a .generate(prompt: str) -> str method.
+        We will wire the real model later.
+        """
+        prompt = build_solver_prompt(question=question, context=context)
+
+        if self.llm_client is None:
+            # Safe placeholder so the app can run while we build the stack.
+            # We will replace this with a real model client later.
+            return "PLACEHOLDER"
+
+        try:
+            return self.llm_client.generate(prompt)
+        except Exception:
+            return "PLACEHOLDER"

app.py

@@ -1,107 +1,128 @@
 import os
 import gradio as gr
 import requests
-import inspect
 import pandas as pd
 
-# (Keep Constants as is)
-# --- Constants ---
+from agent import SubmissionAgent
+
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 
-# --- Basic Agent Definition ---
-# ----- THIS IS WERE YOU CAN BUILD WHAT YOU WANT ------
-class BasicAgent:
-    def __init__(self):
-        print("BasicAgent initialized.")
-    def __call__(self, question: str) -> str:
-        print(f"Agent received question (first 50 chars): {question[:50]}...")
-        fixed_answer = "This is a default answer."
-        print(f"Agent returning fixed answer: {fixed_answer}")
-        return fixed_answer
-
-def run_and_submit_all( profile: gr.OAuthProfile | None):
+
+def run_and_submit_all(profile: gr.OAuthProfile | None):
     """
-    Fetches all questions, runs the BasicAgent on them, submits all answers,
-    and displays the results.
+    Fetch all questions, run the agent on them, submit answers,
+    and display the final score plus a results table.
     """
-    # --- Determine HF Space Runtime URL and Repo URL ---
-    space_id = os.getenv("SPACE_ID") # Get the SPACE_ID for sending link to the code
+    space_id = os.getenv("SPACE_ID")
 
     if profile:
-        username= f"{profile.username}"
+        username = profile.username
         print(f"User logged in: {username}")
     else:
         print("User not logged in.")
-        return "Please Login to Hugging Face with the button.", None
+        return "Please login to Hugging Face first.", None
 
     api_url = DEFAULT_API_URL
     questions_url = f"{api_url}/questions"
     submit_url = f"{api_url}/submit"
 
-    # 1. Instantiate Agent ( modify this part to create your agent)
+    # Instantiate your real agent
     try:
-        agent = BasicAgent()
+        agent = SubmissionAgent()
     except Exception as e:
-        print(f"Error instantiating agent: {e}")
+        print(f"Error initializing agent: {e}")
         return f"Error initializing agent: {e}", None
-    # In the case of an app running as a hugging Face space, this link points toward your codebase ( usefull for others so please keep it public)
-    agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"
-    print(agent_code)
 
-    # 2. Fetch Questions
+    # Public code link required by the benchmark
+    if space_id:
+        agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"
+    else:
+        agent_code = "SPACE_ID_NOT_AVAILABLE"
+
+    print(f"Agent code URL: {agent_code}")
+
+    # Fetch questions
     print(f"Fetching questions from: {questions_url}")
     try:
-        response = requests.get(questions_url, timeout=15)
+        response = requests.get(questions_url, timeout=20)
         response.raise_for_status()
         questions_data = response.json()
+
         if not questions_data:
-             print("Fetched questions list is empty.")
-             return "Fetched questions list is empty or invalid format.", None
+            return "Fetched questions list is empty.", None
+
         print(f"Fetched {len(questions_data)} questions.")
+
     except requests.exceptions.RequestException as e:
         print(f"Error fetching questions: {e}")
         return f"Error fetching questions: {e}", None
-    except requests.exceptions.JSONDecodeError as e:
-         print(f"Error decoding JSON response from questions endpoint: {e}")
-         print(f"Response text: {response.text[:500]}")
-         return f"Error decoding server response for questions: {e}", None
+    except ValueError as e:
+        print(f"Error decoding questions JSON: {e}")
+        return f"Error decoding questions JSON: {e}", None
     except Exception as e:
-        print(f"An unexpected error occurred fetching questions: {e}")
-        return f"An unexpected error occurred fetching questions: {e}", None
+        print(f"Unexpected error fetching questions: {e}")
+        return f"Unexpected error fetching questions: {e}", None
 
-    # 3. Run your Agent
+    # Run agent on all questions
     results_log = []
     answers_payload = []
+
     print(f"Running agent on {len(questions_data)} questions...")
+
     for item in questions_data:
         task_id = item.get("task_id")
         question_text = item.get("question")
+
         if not task_id or question_text is None:
-            print(f"Skipping item with missing task_id or question: {item}")
+            print(f"Skipping malformed item: {item}")
             continue
+
         try:
-            submitted_answer = agent(question_text)
-            answers_payload.append({"task_id": task_id, "submitted_answer": submitted_answer})
-            results_log.append({"Task ID": task_id, "Question": question_text, "Submitted Answer": submitted_answer})
+            submitted_answer = agent(question_text, task_id=task_id)
+
+            answers_payload.append(
+                {
+                    "task_id": task_id,
+                    "submitted_answer": submitted_answer,
+                }
+            )
+
+            results_log.append(
+                {
+                    "Task ID": task_id,
+                    "Question": question_text,
+                    "Submitted Answer": submitted_answer,
+                }
+            )
+
         except Exception as e:
-             print(f"Error running agent on task {task_id}: {e}")
-             results_log.append({"Task ID": task_id, "Question": question_text, "Submitted Answer": f"AGENT ERROR: {e}"})
+            print(f"Error on task {task_id}: {e}")
+            results_log.append(
+                {
+                    "Task ID": task_id,
+                    "Question": question_text,
+                    "Submitted Answer": f"AGENT ERROR: {e}",
+                }
+            )
 
     if not answers_payload:
-        print("Agent did not produce any answers to submit.")
-        return "Agent did not produce any answers to submit.", pd.DataFrame(results_log)
+        print("No answers generated.")
+        return "Agent did not generate any answers.", pd.DataFrame(results_log)
 
-    # 4. Prepare Submission 
-    submission_data = {"username": username.strip(), "agent_code": agent_code, "answers": answers_payload}
-    status_update = f"Agent finished. Submitting {len(answers_payload)} answers for user '{username}'..."
-    print(status_update)
+    # Prepare submission payload
+    submission_data = {
+        "username": username.strip(),
+        "agent_code": agent_code,
+        "answers": answers_payload,
+    }
 
-    # 5. Submit
     print(f"Submitting {len(answers_payload)} answers to: {submit_url}")
+
     try:
         response = requests.post(submit_url, json=submission_data, timeout=60)
         response.raise_for_status()
         result_data = response.json()
+
         final_status = (
             f"Submission Successful!\n"
             f"User: {result_data.get('username')}\n"
@@ -109,88 +130,87 @@ def run_and_submit_all( profile: gr.OAuthProfile | None):
             f"({result_data.get('correct_count', '?')}/{result_data.get('total_attempted', '?')} correct)\n"
             f"Message: {result_data.get('message', 'No message received.')}"
         )
+
         print("Submission successful.")
-        results_df = pd.DataFrame(results_log)
-        return final_status, results_df
+        return final_status, pd.DataFrame(results_log)
+
     except requests.exceptions.HTTPError as e:
         error_detail = f"Server responded with status {e.response.status_code}."
         try:
             error_json = e.response.json()
             error_detail += f" Detail: {error_json.get('detail', e.response.text)}"
-        except requests.exceptions.JSONDecodeError:
+        except ValueError:
             error_detail += f" Response: {e.response.text[:500]}"
+
         status_message = f"Submission Failed: {error_detail}"
         print(status_message)
-        results_df = pd.DataFrame(results_log)
-        return status_message, results_df
+        return status_message, pd.DataFrame(results_log)
+
     except requests.exceptions.Timeout:
-        status_message = "Submission Failed: The request timed out."
+        status_message = "Submission Failed: Request timed out."
         print(status_message)
-        results_df = pd.DataFrame(results_log)
-        return status_message, results_df
+        return status_message, pd.DataFrame(results_log)
+
     except requests.exceptions.RequestException as e:
         status_message = f"Submission Failed: Network error - {e}"
         print(status_message)
-        results_df = pd.DataFrame(results_log)
-        return status_message, results_df
+        return status_message, pd.DataFrame(results_log)
+
     except Exception as e:
-        status_message = f"An unexpected error occurred during submission: {e}"
+        status_message = f"Unexpected submission error: {e}"
         print(status_message)
-        results_df = pd.DataFrame(results_log)
-        return status_message, results_df
+        return status_message, pd.DataFrame(results_log)
 
 
-# --- Build Gradio Interface using Blocks ---
 with gr.Blocks() as demo:
-    gr.Markdown("# Basic Agent Evaluation Runner")
+    gr.Markdown("# Hugging Face Unit 4 Agent Evaluation Runner")
     gr.Markdown(
         """
-        **Instructions:**
-
-        1.  Please clone this space, then modify the code to define your agent's logic, the tools, the necessary packages, etc ...
-        2.  Log in to your Hugging Face account using the button below. This uses your HF username for submission.
-        3.  Click 'Run Evaluation & Submit All Answers' to fetch questions, run your agent, submit answers, and see the score.
-
-        ---
-        **Disclaimers:**
-        Once clicking on the "submit button, it can take quite some time ( this is the time for the agent to go through all the questions).
-        This space provides a basic setup and is intentionally sub-optimal to encourage you to develop your own, more robust solution. For instance for the delay process of the submit button, a solution could be to cache the answers and submit in a seperate action or even to answer the questions in async.
+        Log in with your Hugging Face account, run your agent on all benchmark questions,
+        submit the answers, and view the score plus answer log.
         """
     )
 
-    gr.LoginButton()
-
+    login_button = gr.LoginButton()
     run_button = gr.Button("Run Evaluation & Submit All Answers")
 
-    status_output = gr.Textbox(label="Run Status / Submission Result", lines=5, interactive=False)
-    # Removed max_rows=10 from DataFrame constructor
-    results_table = gr.DataFrame(label="Questions and Agent Answers", wrap=True)
+    status_output = gr.Textbox(
+        label="Run Status / Submission Result",
+        lines=6,
+        interactive=False,
+    )
+
+    results_table = gr.DataFrame(
+        label="Questions and Agent Answers",
+        wrap=True,
+    )
 
     run_button.click(
         fn=run_and_submit_all,
-        outputs=[status_output, results_table]
+        inputs=[login_button],
+        outputs=[status_output, results_table],
     )
 
+
 if __name__ == "__main__":
-    print("\n" + "-"*30 + " App Starting " + "-"*30)
-    # Check for SPACE_HOST and SPACE_ID at startup for information
+    print("\n" + "-" * 30 + " App Starting " + "-" * 30)
+
     space_host_startup = os.getenv("SPACE_HOST")
-    space_id_startup = os.getenv("SPACE_ID") # Get SPACE_ID at startup
+    space_id_startup = os.getenv("SPACE_ID")
 
     if space_host_startup:
-        print(f"✅ SPACE_HOST found: {space_host_startup}")
-        print(f"   Runtime URL should be: https://{space_host_startup}.hf.space")
+        print(f"SPACE_HOST: {space_host_startup}")
+        print(f"Runtime URL: https://{space_host_startup}.hf.space")
     else:
-        print("ℹ️  SPACE_HOST environment variable not found (running locally?).")
+        print("SPACE_HOST not found. Probably running locally.")
 
-    if space_id_startup: # Print repo URLs if SPACE_ID is found
-        print(f"✅ SPACE_ID found: {space_id_startup}")
-        print(f"   Repo URL: https://huggingface.co/spaces/{space_id_startup}")
-        print(f"   Repo Tree URL: https://huggingface.co/spaces/{space_id_startup}/tree/main")
+    if space_id_startup:
+        print(f"SPACE_ID: {space_id_startup}")
+        print(f"Repo URL: https://huggingface.co/spaces/{space_id_startup}")
+        print(f"Repo Tree URL: https://huggingface.co/spaces/{space_id_startup}/tree/main")
     else:
-        print("ℹ️  SPACE_ID environment variable not found (running locally?). Repo URL cannot be determined.")
-
-    print("-"*(60 + len(" App Starting ")) + "\n")
+        print("SPACE_ID not found. Probably running locally.")
 
-    print("Launching Gradio Interface for Basic Agent Evaluation...")
-    demo.launch(debug=True, share=False)
+    print("-" * 75 + "\n")
+    print("Launching Gradio app...")
+    demo.launch(debug=True)

prompts.py

@@ -0,0 +1,55 @@
+from __future__ import annotations
+
+
+SYSTEM_PROMPT = """
+You are a benchmark-solving AI agent.
+
+Your task is to answer questions as accurately as possible.
+
+Rules:
+- Return ONLY the final answer.
+- Do NOT include explanations.
+- Do NOT include reasoning.
+- Do NOT include the words "FINAL ANSWER".
+- Do NOT include labels like "Answer:".
+- Output must be exactly the answer text.
+
+Formatting rules:
+- If the answer is a number, output only the number.
+- If the answer is a word or phrase, output only that word or phrase.
+- If the answer is a date, return the exact date string.
+- Do not add punctuation unless it is part of the answer.
+
+Your response must contain only the final answer string.
+"""
+
+
+def build_solver_prompt(question: str, context: str = "") -> str:
+    """
+    Builds the final prompt sent to the model.
+    Includes optional file context when a task provides additional data.
+    """
+
+    if context:
+        prompt = f"""
+{SYSTEM_PROMPT}
+
+Context information:
+{context}
+
+Question:
+{question}
+
+Return only the final answer.
+"""
+    else:
+        prompt = f"""
+{SYSTEM_PROMPT}
+
+Question:
+{question}
+
+Return only the final answer.
+"""
+
+    return prompt.strip()

tools.py

@@ -0,0 +1,182 @@
+from __future__ import annotations
+import io
+import json
+import os
+from pathlib import Path
+from typing import Optional
+import pandas as pd
+import requests
+
+class TaskFileTool:
+    """
+    Downloads and reads task-linked files from the Hugging Face
+    Unit 4 scoring API.
+
+    Supported text extration:
+    - txt
+    - csv
+    - json
+    - md
+    - html
+    - xml
+
+    For unsupported or binary files, it safely returns an empty string for now.
+    We can extend this later for PDF/images if needed.
+    """
+
+    def __init__(self, api_base_url: str, cache_dir:str = "task_files", timeout: int =30):
+        self.api_base_url = api_base_url.strip("/")
+        self.cache_dir = Path(cache_dir)
+        self.cache_dir.mkdir(parents=True, exist_ok=True)
+        self.timeout = timeout
+    
+    def get_task_context(self, task_id: str) -> str:
+        """
+        Main entry point used by the agent:
+        1. download the task file if present
+        2. read it into text context if supported
+        """
+        file_path = self.download_task_file(task_id)
+        if file_path is None:
+            return ""
+        return self.read_file_as_text(file_path)
+
+    def download_task_file(self, task_id: str) -> Optional[Path]:
+        """
+        Downloads the file linked to a task_id using:
+        GET /files/{task_id}
+
+        Returns:
+            Path to saved file if successful, else None
+        """
+        url = f"{self.api_base_url}/file/{task_id}"
+
+        try:
+            response = requests.get(url, timeout=self.timeout)
+        except requests.RequestException:
+            return None
+        
+        if response.status_code !=200:
+            return None
+        
+        filename = self._infer_filename(response=response, task_id=task_id)
+        file_path = self.cache_dir / filename
+
+        try:
+            with open(file_path, "wb") as f:
+                f.write(response.content)
+            return file_path
+        except OSError:
+            return None
+        return file_path
+    
+    def read_file_as_text(self, file_path: Path) -> str:
+        """
+        Reads supported file types into plain text.
+        """
+        suffix = file_path.suffix.lower()
+
+        try:
+            if suffix in {".txt", ".md", ".html", ".xml", ".csv", ".json"}:
+                return self._read_supported_text_file(file_path, suffix)
+
+            # common fallback for files saved without extension but actually text
+            if suffix == "":
+                return self._read_extensionless_file(file_path)
+
+            return ""
+        except Exception:
+            return ""
+
+    def _read_supported_text_file(self, file_path: Path, suffix: str) -> str:
+        if suffix in {".txt", ".md", ".html", ".xml"}:
+            return file_path.read_text(encoding="utf-8", errors="ignore")
+
+        if suffix == ".json":
+            raw = file_path.read_text(encoding="utf-8", errors="ignore")
+            try:
+                parsed = json.loads(raw)
+                return json.dumps(parsed, indent=2, ensure_ascii=False)
+            except json.JSONDecodeError:
+                return raw
+
+        if suffix == ".csv":
+            try:
+                df = pd.read_csv(file_path)
+                return df.to_csv(index=False)
+            except Exception:
+                return file_path.read_text(encoding="utf-8", errors="ignore")
+
+        return ""
+
+    def _read_extensionless_file(self, file_path: Path) -> str:
+        """
+        Try to interpret extensionless files as utf-8 text first.
+        """
+        try:
+            raw = file_path.read_text(encoding="utf-8", errors="ignore")
+            if raw.strip():
+                return raw
+        except Exception:
+            pass
+        return ""
+
+    def _infer_filename(self, response: requests.Response, task_id: str) -> str:
+        """
+        Attempts to infer a useful filename from headers.
+        Falls back to task_id if no filename is available.
+        """
+        content_disposition = response.headers.get("content-disposition", "")
+        filename = self._extract_filename_from_content_disposition(content_disposition)
+
+        if filename:
+            return self._safe_filename(filename)
+
+        content_type = response.headers.get("content-type", "").lower()
+        extension = self._extension_from_content_type(content_type)
+
+        if extension:
+            return f"{task_id}{extension}"
+
+        return str(task_id)
+
+    @staticmethod
+    def _extract_filename_from_content_disposition(content_disposition: str) -> Optional[str]:
+        """
+        Example header:
+        content-disposition: attachment; filename="example.csv"
+        """
+        if "filename=" not in content_disposition:
+            return None
+
+        try:
+            filename = content_disposition.split("filename=")[-1].strip().strip('"')
+            return filename or None
+        except Exception:
+            return None
+
+    @staticmethod
+    def _extension_from_content_type(content_type: str) -> str:
+        mapping = {
+            "text/plain": ".txt",
+            "text/csv": ".csv",
+            "application/csv": ".csv",
+            "application/json": ".json",
+            "text/markdown": ".md",
+            "text/html": ".html",
+            "application/xml": ".xml",
+            "text/xml": ".xml",
+        }
+
+        for key, ext in mapping.items():
+            if key in content_type:
+                return ext
+
+        return ""
+
+    @staticmethod
+    def _safe_filename(filename: str) -> str:
+        """
+        Prevent path traversal and weird path issues.
+        """
+        return os.path.basename(filename)

utils.py

@@ -0,0 +1,100 @@
+from __future__ import annotations
+
+import re
+
+
+def extract_final_answer(text: str) -> str:
+    """
+    Extract the most likely final answer from raw model output.
+
+    In V1 we keep this conservative:
+    - if the model returns a normal short answer, keep it
+    - if it adds common prefixes like 'Answer:' or 'Final answer:', remove them
+    - if it returns multiple lines, prefer the last non-empty line
+    """
+    if text is None:
+        return ""
+
+    text = str(text).strip()
+    if not text:
+        return ""
+
+    # Remove fenced code blocks if the model wraps the answer oddly
+    text = re.sub(r"^```[a-zA-Z0-9_-]*\s*", "", text)
+    text = re.sub(r"\s*```$", "", text)
+
+    # Common exact-answer markers
+    marker_patterns = [
+        r"(?i)\bfinal answer\s*:\s*",
+        r"(?i)\banswer\s*:\s*",
+        r"(?i)\bthe answer is\s*:\s*",
+        r"(?i)\bthe answer is\s+",
+    ]
+
+    cleaned = text
+    for pattern in marker_patterns:
+        cleaned = re.sub(pattern, "", cleaned).strip()
+
+    # If multi-line, prefer the last meaningful line
+    lines = [line.strip() for line in cleaned.splitlines() if line.strip()]
+    if not lines:
+        return ""
+
+    if len(lines) == 1:
+        return lines[0]
+
+    return lines[-1]
+
+
+def normalize_final_answer(text: str) -> str:
+    """
+    Normalize answer text for safer exact-match submission without being too aggressive.
+
+    Rules:
+    - trim outer whitespace
+    - collapse internal repeated whitespace
+    - remove wrapping quotes if they wrap the full answer
+    - remove a single trailing period only for plain word/phrase answers
+      but keep decimal numbers and date punctuation intact
+    """
+    if text is None:
+        return ""
+
+    text = str(text).strip()
+    if not text:
+        return ""
+
+    # Collapse repeated whitespace
+    text = re.sub(r"\s+", " ", text).strip()
+
+    # Remove matching surrounding quotes
+    if len(text) >= 2:
+        if (text[0] == text[-1]) and text[0] in {'"', "'"}:
+            text = text[1:-1].strip()
+
+    # Remove common leading labels again, just in case
+    text = re.sub(r"(?i)^(final answer|answer)\s*:\s*", "", text).strip()
+
+    # Remove one trailing period for simple phrase answers only
+    # Keep decimals like 3.14 intact
+    if text.endswith("."):
+        if not re.fullmatch(r"\d+\.\d+", text):
+            text = text[:-1].strip()
+
+    return text
+
+
+def is_placeholder_answer(text: str) -> bool:
+    """
+    Detect placeholder/fallback outputs so app.py can optionally flag them.
+    """
+    if text is None:
+        return True
+
+    normalized = normalize_final_answer(text).lower()
+    return normalized in {
+        "",
+        "placeholder",
+        "n/a",
+        "unknown",
+    }