Add agent tools

definitions.py

@@ -0,0 +1,15 @@
+# pylint: disable=C0115
+from typing import TypedDict, Literal
+
+
+class TranscriptionFile(TypedDict):
+    TranscriptFileUri: str
+
+
+class TranscriptionResponse(TypedDict):
+    TranscriptionJobStatus: Literal["IN_PROGRESS", "COMPLETED", "FAILED"]
+    Transcript: TranscriptionFile
+
+
+class TranscriptionJob(TypedDict):
+    TranscriptionJob: TranscriptionResponse

requirements.txt

@@ -1,2 +1,76 @@
-gradio
-requests
+aiofiles==24.1.0 ; python_version >= '3.8'
+annotated-types==0.7.0 ; python_version >= '3.8'
+anyio==4.9.0 ; python_version >= '3.9'
+boto3==1.38.23
+botocore==1.38.23 ; python_version >= '3.9'
+certifi==2025.4.26 ; python_version >= '3.6'
+charset-normalizer==3.4.2 ; python_version >= '3.7'
+click==8.2.1 ; python_version >= '3.10'
+et-xmlfile==2.0.0 ; python_version >= '3.8'
+fastapi==0.115.12 ; python_version >= '3.8'
+ffmpy==0.5.0 ; python_version >= '3.8' and python_version < '4.0'
+filelock==3.18.0 ; python_version >= '3.9'
+fsspec==2025.5.1 ; python_version >= '3.9'
+gradio==5.31.0
+gradio-client==1.10.1 ; python_version >= '3.10'
+groovy==0.1.2 ; python_version >= '3.10'
+h11==0.16.0 ; python_version >= '3.8'
+hf-xet==1.1.2 ; platform_machine == 'x86_64' or platform_machine == 'amd64' or platform_machine == 'arm64' or platform_machine == 'aarch64'
+httpcore==1.0.9 ; python_version >= '3.8'
+httpx==0.28.1 ; python_version >= '3.8'
+huggingface-hub==0.32.1 ; python_full_version >= '3.8.0'
+idna==3.10 ; python_version >= '3.6'
+jinja2==3.1.6 ; python_version >= '3.7'
+jmespath==1.0.1 ; python_version >= '3.7'
+jsonpatch==1.33 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4, 3.5, 3.6'
+jsonpointer==3.0.0 ; python_version >= '3.7'
+langchain-core==0.3.61 ; python_version >= '3.9'
+langgraph==0.4.7
+langgraph-checkpoint==2.0.26 ; python_version >= '3.9'
+langgraph-prebuilt==0.2.1 ; python_version >= '3.9'
+langgraph-sdk==0.1.70 ; python_version >= '3.9'
+langsmith==0.3.42 ; python_version >= '3.9'
+markdown-it-py==3.0.0 ; python_version >= '3.8'
+markupsafe==3.0.2 ; python_version >= '3.9'
+mdurl==0.1.2 ; python_version >= '3.7'
+numpy==2.2.6 ; python_version >= '3.10'
+openpyxl==3.1.5
+orjson==3.10.18 ; python_version >= '3.9'
+ormsgpack==1.10.0 ; python_version >= '3.9'
+packaging==24.2 ; python_version >= '3.8'
+pandas==2.2.3 ; python_version >= '3.9'
+pillow==11.2.1 ; python_version >= '3.9'
+pydantic==2.11.5 ; python_version >= '3.9'
+pydantic-core==2.33.2 ; python_version >= '3.9'
+pydub==0.25.1
+pygments==2.19.1 ; python_version >= '3.8'
+pymupdf==1.26.0
+python-dateutil==2.9.0.post0 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
+python-dotenv==1.1.0
+python-multipart==0.0.20 ; python_version >= '3.8'
+pytz==2025.2
+pyyaml==6.0.2 ; python_version >= '3.8'
+requests==2.32.3
+requests-toolbelt==1.0.0 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
+rich==14.0.0 ; python_full_version >= '3.8.0'
+ruff==0.11.11 ; sys_platform != 'emscripten'
+s3transfer==0.13.0 ; python_version >= '3.9'
+safehttpx==0.1.6 ; python_version >= '3.10'
+semantic-version==2.10.0 ; python_version >= '2.7'
+shellingham==1.5.4 ; python_version >= '3.7'
+six==1.17.0 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
+smolagents==1.16.1
+sniffio==1.3.1 ; python_version >= '3.7'
+starlette==0.46.2 ; sys_platform != 'emscripten'
+tenacity==9.1.2 ; python_version >= '3.9'
+tomlkit==0.13.2 ; python_version >= '3.8'
+tqdm==4.67.1 ; python_version >= '3.7'
+typer==0.16.0 ; sys_platform != 'emscripten'
+typing-extensions==4.13.2 ; python_version >= '3.8'
+typing-inspection==0.4.1 ; python_version >= '3.9'
+tzdata==2025.2 ; python_version >= '2'
+urllib3==2.4.0 ; python_version >= '3.9'
+uvicorn==0.34.2 ; sys_platform != 'emscripten'
+websockets==15.0.1 ; python_version >= '3.9'
+xxhash==3.5.0 ; python_version >= '3.7'
+zstandard==0.23.0 ; python_version >= '3.8'

tools.py

@@ -0,0 +1,139 @@
+# pylint: disable=W0718
+import ast
+import json
+import os
+from time import sleep
+from uuid import uuid4
+
+import boto3
+import fitz
+from pandas import read_excel
+from smolagents import tool, Tool
+
+from definitions import TranscriptionJob
+from utils import get_file, s3_upload_file, s3_download_file
+
+
+@tool
+def math_calculator(query: str) -> str:
+    """A simple calculator tool that evaluates mathematical expressions.
+
+    Args:
+        query (str): A mathematical expression as a string, e.g., "2 + 2 * 3".
+    """
+    try:
+        result = ast.literal_eval(query)
+        return str(result)
+    except Exception as e:
+        return f"Error evaluating expression: {e}"
+
+
+@tool
+def excel_reader(task_id: str, file_name: str) -> str:
+    """Reads an Excel file and returns its content as a dataframe string.
+
+    Args:
+        task_id (str): The ID of the task associated with the file.
+        file_name (str): The name of the Excel file to read.
+    """
+    try:
+        file_content = get_file(task_id)
+        df = read_excel(file_content, engine="openpyxl")
+        return df.to_string(index=False)
+    except Exception as e:
+        return f"Error reading Excel file {file_name}: {e}"
+
+
+def txt_reader(task_id: str, file_name: str) -> str:
+    """Reads a text file and returns its content as a string.
+
+    Args:
+        task_id (str): The ID of the task associated with the file.
+        file_name (str): The name of the file to read.
+    """
+    try:
+        file_content = get_file(task_id)
+        return file_content.read().decode("utf-8")
+    except Exception as e:
+        return f"Error reading file {file_name}: {e}"
+
+
+def pdf_reader(task_id: str, file_name: str) -> str:
+    """Reads a PDF file and returns its content as a string.
+
+    Args:
+        task_id (str): The ID of the task associated with the file.
+        file_name (str): The name of the PDF file to read.
+    """
+    try:
+        file_content = get_file(task_id)
+        with fitz.open(stream=file_content.read(), filetype="pdf") as doc:
+            content = [page.get_text() for page in doc if page.get_text()]
+            text = "\n".join(content)
+            if not text:
+                return f"No text found in PDF file {file_name}."
+            return text.strip()
+    except Exception as e:
+        return f"Error reading PDF file {file_name}: {e}"
+
+
+class AudioTranscriber(Tool):  # pylint: disable=C0115
+    name = "AudioTranscriber"
+    description = "Extract text from audio files, such as MP3, MP4, WAV, etc."
+    inputs = {
+        "task_id": {
+            "type": "string",
+            "description": "The ID of the task associated with the audio file.",
+        },
+        "file_name": {
+            "type": "string",
+            "description": "The name of the audio file to transcribe.",
+        },
+    }
+
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+        region = os.getenv("AWS_REGION", "us-east-1")
+        self.client = boto3.client("transcribe", region_name=region)
+
+    def _transcribe_audio(self, job_name: str, media_uri: str) -> dict:
+        self.client.start_transcription_job(
+            TranscriptionJobName=job_name,
+            Media={"MediaFileUri": media_uri},
+            IdentifyLanguage=True,
+            OutputBucketName=os.getenv("TARGET_BUCKET"),
+            OutputKey=f"{job_name}.json",
+        )
+
+    def _get_transcription(self, job_name: str) -> str:
+        while True:
+            response: TranscriptionJob = self.client.get_transcription_job(TranscriptionJobName=job_name)
+            status = response["TranscriptionJob"]["TranscriptionJobStatus"]
+            if status in ["COMPLETED", "FAILED"]:
+                break
+            sleep(5)
+
+        transcript_url = response["TranscriptionJob"]["Transcript"]["TranscriptFileUri"]
+        try:
+            bytes_result = s3_download_file(os.getenv("TARGET_BUCKET"), transcript_url.split("/")[-1])
+            transcription_data = json.loads(bytes_result.read().decode("utf-8"))
+            return transcription_data["transcripts"][0]["transcript"]
+        except json.JSONDecodeError as e:
+            print(f"Error decoding transcription JSON: {e}")
+            raise
+        except Exception as e:
+            print(f"Error downloading or processing transcription file: {e}")
+            raise
+
+    def forward(self, task_id: str, file_name: str) -> str:  # pylint: disable=W0221
+        try:
+            file_content = get_file(task_id)
+            s3_upload_file(file_content, os.getenv("SOURCE_BUCKET"), file_name)
+
+            media_uri = f"s3://{os.getenv('SOURCE_BUCKET')}/{file_name}"
+            job_name = f"{uuid4()}-{file_name.split('.')[0]}"
+            self._transcribe_audio(job_name, media_uri)
+            transcription = self._get_transcription(job_name)
+            return transcription
+        except Exception as e:
+            return f"Error starting transcription job for {file_name}: {e}"

utils.py

@@ -0,0 +1,109 @@
+import os
+from io import BytesIO
+
+import boto3
+import requests
+from dotenv import load_dotenv
+
+DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
+
+
+load_dotenv()
+
+bedrock_client = boto3.client(
+    "bedrock-runtime",
+    region_name=os.getenv("AWS_REGION"),
+    aws_access_key_id=os.getenv("AWS_ACCESS_KEY_ID"),
+    aws_secret_access_key=os.getenv("AWS_SECRET_ACCESS_KEY"),
+)
+
+
+def get_file(task_id: str) -> BytesIO:
+    """Fetches a file associated with a task ID from the default API URL.
+
+    Parameters
+    ----------
+    task_id : str
+        The ID of the task for which the file is to be fetched.
+
+    Returns
+    -------
+    BytesIO
+        A BytesIO object containing the file content.
+    Raises
+    ------
+    requests.exceptions.RequestException
+        If there is an error during the HTTP request.
+    Exception
+        For any other unexpected errors that may occur.
+    """
+    url = f"{DEFAULT_API_URL}/files/{task_id}"
+    try:
+        response = requests.get(url, timeout=15)
+        response.raise_for_status()
+        return BytesIO(response.content)
+    except requests.exceptions.RequestException as e:
+        print(f"Error fetching file for task {task_id}: {e}")
+        raise
+    except Exception as e:
+        print(f"An unexpected error occurred fetching file for task {task_id}: {e}")
+        raise
+
+
+def s3_upload_file(file_content: BytesIO, bucket_name: str, object_name: str) -> None:
+    """Uploads a file to an S3 bucket.
+
+    Parameters
+    ----------
+    file_content : BytesIO
+        The content of the file to upload.
+    bucket_name : str
+        The name of the S3 bucket.
+    object_name : str
+        The name of the object in the S3 bucket.
+
+    Raises
+    ------
+    Exception
+        If there is an error during the upload process.
+    """
+    try:
+        s3_client = boto3.client("s3")
+        s3_client.put_object(
+            Bucket=bucket_name,
+            Key=object_name,
+            Body=file_content.getvalue(),
+            ContentType="application/octet-stream",
+        )
+    except Exception as e:
+        print(f"Error uploading file to S3: {e}")
+        raise
+
+
+def s3_download_file(bucket_name: str, object_name: str) -> BytesIO:
+    """Downloads a file from an S3 bucket.
+
+    Parameters
+    ----------
+    bucket_name : str
+        The name of the S3 bucket.
+    object_name : str
+        The name of the object in the S3 bucket.
+
+    Returns
+    -------
+    BytesIO
+        A BytesIO object containing the downloaded file content.
+
+    Raises
+    ------
+    Exception
+        If there is an error during the download process.
+    """
+    try:
+        s3_client = boto3.client("s3")
+        response = s3_client.get_object(Bucket=bucket_name, Key=object_name)
+        return BytesIO(response["Body"].read())
+    except Exception as e:
+        print(f"Error downloading file from S3: {e}")
+        raise