Initial commit with RAG, FastAPI and Gradio UI

.gitignore

@@ -0,0 +1,25 @@
+# Virtual Environments
+venv/
+env/
+.venv/
+
+# Environment Variables
+.env
+*.env
+
+# Python Cache
+__pycache__/
+*.pyc
+*.pyo
+*.pyd
+
+# ChromaDB local vector storage
+chroma_db/
+
+# IDEs
+.vscode/
+.idea/
+
+# OS generated files
+.DS_Store
+Thumbs.db

README.md

@@ -0,0 +1,54 @@
+# Task Prompting Tool
+
+A tool designed for project managers and team leaders to generate developer-ready prompts. By providing a task description, specifying a field (e.g., Backend, Frontend), and uploading relevant context files, the underlying LLM (OpenAI or Google Gemini) will construct a comprehensive prompt ready to be handed to developers.
+
+## Prerequisites
+
+- Python 3.8+
+- API keys for OpenAI and/or Google Gen AI.
+
+## Setup Instructions
+
+**1. Create Hand-configured Keys in `.env`**
+Edit the `.env` file and insert your API keys and models as preferred:
+```env
+OPENAI_API_KEY="your-openai-api-key"
+GOOGLE_API_KEY="your-google-api-key"
+LLM_PROVIDER="google" # Options: "google" or "openai"
+OPENAI_MODEL="gpt-4.1-mini"
+GOOGLE_MODEL="gemini-3.1-flash-lite-preview"
+```
+
+**2. Setup Virtual Environment & Install Dependencies**
+Open a terminal and run the following commands in the project directory:
+
+```bash
+python3 -m venv venv
+source venv/bin/activate
+pip install -r requirements.txt
+```
+
+*(On Windows, activate the virtual environment using `venv\Scripts\activate`)*
+
+## Running the Application
+
+This project features both a FastAPI backend (providing an API) and a Gradio frontend (providing a UI).
+
+### Method 1: Using the UI (Recommended)
+You can directly run the Gradio application to access the user interface.
+
+```bash
+source venv/bin/activate
+python gradio_app.py
+```
+After executing, an interface will open at `http://127.0.0.1:7860/` by default. You can open your browser to this URL to interact with the Task Prompting Tool.
+
+### Method 2: Running the API Server
+If you'd like to integrate this logic into another system or frontend, run the FastAPI backend:
+
+```bash
+source venv/bin/activate
+python api.py
+```
+The server will run at `http://0.0.0.0:8000`. 
+- View API Documentation at `http://127.0.0.1:8000/docs` to test endpoints interactively.

agent.py

@@ -0,0 +1,107 @@
+import os
+from dotenv import load_dotenv
+from langchain_openai import ChatOpenAI, OpenAIEmbeddings
+from langchain_google_genai import ChatGoogleGenerativeAI, GoogleGenerativeAIEmbeddings
+from langchain_core.prompts import ChatPromptTemplate
+from langchain_core.output_parsers import StrOutputParser
+from langchain_core.documents import Document
+from langchain_community.vectorstores import Chroma
+from langchain_text_splitters import RecursiveCharacterTextSplitter
+
+load_dotenv()
+
+def get_llm():
+    provider = os.getenv("LLM_PROVIDER", "google").lower()
+    
+    if provider == "openai":
+        return ChatOpenAI(
+            model=os.getenv("OPENAI_MODEL", "gpt-4.1-mini"), 
+            temperature=0.7
+        )
+    elif provider == "google":
+        return ChatGoogleGenerativeAI(
+            model=os.getenv("GOOGLE_MODEL", "gemini-3.1-flash-lite-preview"),
+            temperature=0.7
+        )
+    else:
+        raise ValueError(f"Unknown LLM Provider: {provider}")
+
+def get_embeddings():
+    provider = os.getenv("LLM_PROVIDER", "google").lower()
+    if provider == "openai":
+        return OpenAIEmbeddings(model="text-embedding-3-small")
+    elif provider == "google":
+        return GoogleGenerativeAIEmbeddings(model="models/gemini-embedding-001")
+    else:
+        raise ValueError(f"Unknown LLM Provider: {provider}")
+
+def process_and_retrieve_context(description: str, field: str, files_data: list[dict]) -> str:
+    """Takes a list of file dictionaries and retrieves relevant context using ChromaDB."""
+    if not files_data:
+        return "No extra files provided."
+        
+    docs = []
+    for file in files_data:
+        docs.append(Document(
+            page_content=file["content"],
+            metadata={"source": file["filename"]}
+        ))
+        
+    # Split the documents
+    text_splitter = RecursiveCharacterTextSplitter(
+        chunk_size=1000, 
+        chunk_overlap=200
+    )
+    splits = text_splitter.split_documents(docs)
+    
+    # Store locally in chromadb directory and use it to retrieve
+    vectorstore = Chroma.from_documents(
+        documents=splits, 
+        embedding=get_embeddings(),
+        persist_directory="./chroma_db"
+    )
+    
+    # Use description and field to retrieve relevant chunks
+    query = f"Field: {field}. Task: {description}"
+    retriever = vectorstore.as_retriever(search_kwargs={"k": 5})
+    
+    retrieved_docs = retriever.invoke(query)
+    
+    context = ""
+    for idx, doc in enumerate(retrieved_docs):
+        context += f"\n--- Retrieved Chunk {idx+1} from {doc.metadata.get('source', 'Unknown')} ---\n{doc.page_content}\n"
+        
+    return context
+
+def generate_task_prompt(description: str, field: str, files_data: list[dict]) -> str:
+    llm = get_llm()
+    
+    # Get filtered context via RAG
+    files_context = process_and_retrieve_context(description, field, files_data)
+    
+    system_prompt = (
+        "You are an expert technical project manager and architect. "
+        "Your goal is to take a task description provided by a project manager, context about the field (e.g., backend, frontend), "
+        "and any uploaded file context, and produce a high-quality, developer-ready task prompt.\n\n"
+        "Return ONLY the finalized prompt ready to be handed to a developer."
+    )
+    
+    human_prompt = (
+        "Field/Domain: {field}\n"
+        "Task Description:\n{description}\n\n"
+        "Relevant Code/Files Context:\n{files_context}\n\n"
+        "Please generate a comprehensive developer prompt."
+    )
+    
+    prompt = ChatPromptTemplate.from_messages([
+        ("system", system_prompt),
+        ("human", human_prompt),
+    ])
+    
+    chain = prompt | llm | StrOutputParser()
+    
+    return chain.invoke({
+        "field": field,
+        "description": description,
+        "files_context": files_context
+    })

api.py

@@ -0,0 +1,47 @@
+from fastapi import FastAPI, UploadFile, File, Form
+from typing import List, Optional
+import uvicorn
+from agent import generate_task_prompt
+
+import io
+import pypdf
+import docx2txt
+
+app = FastAPI(title="Task Prompting API")
+
+@app.post("/generate_prompt")
+async def generate_prompt(
+    description: str = Form(...),
+    field: str = Form(...),
+    files: Optional[List[UploadFile]] = File(None)
+):
+    files_data = []
+    
+    if files:
+        for file in files:
+            content = await file.read()
+            filename_lower = file.filename.lower()
+            
+            try:
+                if filename_lower.endswith(".pdf"):
+                    reader = pypdf.PdfReader(io.BytesIO(content))
+                    text = "\n".join([page.extract_text() or "" for page in reader.pages])
+                    files_data.append({"filename": file.filename, "content": text})
+                elif filename_lower.endswith(".docx"):
+                    text = docx2txt.process(io.BytesIO(content))
+                    files_data.append({"filename": file.filename, "content": text})
+                else:
+                    decoded_content = content.decode('utf-8')
+                    files_data.append({"filename": file.filename, "content": decoded_content})
+            except Exception as e:
+                print(f"Error processing {file.filename}: {e}")
+                pass # Skip files that fail encoding or extraction
+                
+    try:
+        generated_prompt = generate_task_prompt(description, field, files_data)
+        return {"status": "success", "prompt": generated_prompt}
+    except Exception as e:
+        return {"status": "error", "message": str(e)}
+
+if __name__ == "__main__":
+    uvicorn.run("api:app", host="0.0.0.0", port=8000, reload=True)

gradio_app.py

@@ -0,0 +1,71 @@
+import gradio as gr
+import requests
+import os
+
+from agent import generate_task_prompt
+
+import pypdf
+import docx2txt
+
+def process_request(description, field, uploaded_files):
+    files_data = []
+    
+    if uploaded_files:
+        for file_path in uploaded_files:
+            filename = os.path.basename(file_path)
+            lower_name = filename.lower()
+            try:
+                if lower_name.endswith(".pdf"):
+                    reader = pypdf.PdfReader(file_path)
+                    content = "\n".join([page.extract_text() or "" for page in reader.pages])
+                elif lower_name.endswith(".docx"):
+                    content = docx2txt.process(file_path)
+                else:
+                    with open(file_path, "r", encoding="utf-8") as f:
+                        content = f.read()
+                files_data.append({"filename": filename, "content": content})
+            except Exception as e:
+                print(f"Skipping binary/unreadable file: {filename}")
+                
+    try:
+        return generate_task_prompt(description, field, files_data)
+    except Exception as e:
+        return f"Error occurred: {str(e)}"
+
+# Gradio Interface
+with gr.Blocks(title="Task Prompting Tool") as demo:
+    gr.Markdown("# 🚀 Developer Task Prompting Tool (RAG Enabled)")
+    gr.Markdown("Generate high-quality, developer-ready prompts using advanced ChromaDB chunking for very large projects.")
+    
+    with gr.Row():
+        with gr.Column(scale=2):
+            field_input = gr.Dropdown(
+                choices=["Backend", "Frontend", "Fullstack", "DevOps", "Data Science", "Mobile", "Other"], 
+                label="Field / Application Area",
+                value="Backend"
+            )
+            desc_input = gr.Textbox(
+                label="Task Description", 
+                placeholder="Describe what needs to be done...", 
+                lines=5
+            )
+            file_input = gr.File(
+                label="Upload Context Files (Code, MD, JSON, etc.)", 
+                file_count="multiple"
+            )
+            submit_btn = gr.Button("Generate Prompt", variant="primary")
+            
+        with gr.Column(scale=3):
+            output_text = gr.Textbox(
+                label="Generated Task Prompt", 
+                lines=15
+            )
+            
+    submit_btn.click(
+        fn=process_request,
+        inputs=[desc_input, field_input, file_input],
+        outputs=output_text
+    )
+
+if __name__ == "__main__":
+    demo.launch(server_name="0.0.0.0", server_port=7860)

requirements.txt

@@ -0,0 +1,147 @@
+aiofiles==24.1.0
+aiohappyeyeballs==2.6.1
+aiohttp==3.13.5
+aiosignal==1.4.0
+annotated-doc==0.0.4
+annotated-types==0.7.0
+anyio==4.13.0
+attrs==26.1.0
+bcrypt==5.0.0
+beautifulsoup4==4.14.3
+brotli==1.2.0
+bs4==0.0.2
+build==1.4.2
+certifi==2026.2.25
+cffi==2.0.0
+charset-normalizer==3.4.7
+chromadb==1.5.5
+click==8.3.2
+cryptography==46.0.6
+dataclasses-json==0.6.7
+distro==1.9.0
+docx2txt==0.9
+durationpy==0.10
+fastapi==0.135.3
+ffmpy==1.0.0
+filelock==3.25.2
+filetype==1.2.0
+flatbuffers==25.12.19
+frozenlist==1.8.0
+fsspec==2026.3.0
+google-auth==2.49.1
+google-genai==1.70.0
+googleapis-common-protos==1.74.0
+gradio==6.11.0
+gradio_client==2.4.0
+greenlet==3.3.2
+groovy==0.1.2
+grpcio==1.80.0
+h11==0.16.0
+hf-gradio==0.3.0
+hf-xet==1.4.3
+httpcore==1.0.9
+httptools==0.7.1
+httpx==0.28.1
+httpx-sse==0.4.3
+huggingface_hub==1.9.0
+idna==3.11
+importlib_metadata==8.7.1
+importlib_resources==6.5.2
+Jinja2==3.1.6
+jiter==0.13.0
+jsonpatch==1.33
+jsonpointer==3.1.1
+jsonschema==4.26.0
+jsonschema-specifications==2025.9.1
+kubernetes==35.0.0
+langchain==1.2.15
+langchain-classic==1.0.3
+langchain-community==0.4.1
+langchain-core==1.2.25
+langchain-google-genai==4.2.1
+langchain-openai==1.1.12
+langchain-text-splitters==1.1.1
+langgraph==1.1.6
+langgraph-checkpoint==4.0.1
+langgraph-prebuilt==1.0.9
+langgraph-sdk==0.3.12
+langsmith==0.7.25
+markdown-it-py==4.0.0
+MarkupSafe==3.0.3
+marshmallow==3.26.2
+mdurl==0.1.2
+mmh3==5.2.1
+mpmath==1.3.0
+multidict==6.7.1
+mypy_extensions==1.1.0
+numpy==2.4.4
+oauthlib==3.3.1
+onnxruntime==1.24.4
+openai==2.30.0
+opentelemetry-api==1.40.0
+opentelemetry-exporter-otlp-proto-common==1.40.0
+opentelemetry-exporter-otlp-proto-grpc==1.40.0
+opentelemetry-proto==1.40.0
+opentelemetry-sdk==1.40.0
+opentelemetry-semantic-conventions==0.61b0
+orjson==3.11.8
+ormsgpack==1.12.2
+overrides==7.7.0
+packaging==26.0
+pandas==3.0.2
+pillow==12.2.0
+propcache==0.4.1
+protobuf==6.33.6
+pyasn1==0.6.3
+pyasn1_modules==0.4.2
+pybase64==1.4.3
+pycparser==3.0
+pydantic==2.12.5
+pydantic-settings==2.13.1
+pydantic_core==2.41.5
+pydub==0.25.1
+Pygments==2.20.0
+pypdf==6.9.2
+PyPika==0.51.1
+pyproject_hooks==1.2.0
+python-dateutil==2.9.0.post0
+python-dotenv==1.2.2
+python-multipart==0.0.22
+pytz==2026.1.post1
+PyYAML==6.0.3
+referencing==0.37.0
+regex==2026.3.32
+requests==2.33.1
+requests-oauthlib==2.0.0
+requests-toolbelt==1.0.0
+rich==14.3.3
+rpds-py==0.30.0
+safehttpx==0.1.7
+semantic-version==2.10.0
+shellingham==1.5.4
+six==1.17.0
+sniffio==1.3.1
+soupsieve==2.8.3
+SQLAlchemy==2.0.49
+starlette==1.0.0
+sympy==1.14.0
+tenacity==9.1.4
+tiktoken==0.12.0
+tokenizers==0.22.2
+tomlkit==0.13.3
+tqdm==4.67.3
+typer==0.24.1
+typing-inspect==0.9.0
+typing-inspection==0.4.2
+typing_extensions==4.15.0
+urllib3==2.6.3
+uuid_utils==0.14.1
+uvicorn==0.43.0
+uvloop==0.22.1
+watchfiles==1.1.1
+websocket-client==1.9.0
+websockets==16.0
+xxhash==3.6.0
+yarl==1.23.0
+zipp==3.23.0
+zstandard==0.25.0