Aniket2692/SDLC_Agentic_Workflow

End to End Agentic SDLC Project

🧠 Agentic AI for Software Design Life Cycle (SDLC)

An AI-powered pipeline that automates the entire SDLC using modular LLM agents built with LangGraph and LangChain.

🚀 Features

• —✅ User story generation from software requirements
• —📘 Design document sectioning and authoring
• —🧑‍💻 Source code generation (Python)
• —🔍 Code review and feedback loops
• —🔐 Security checks
• —🧪 Test case creation (with pytest)
• —🎯 Human-in-the-loop approvals for each step
• —🌐 Streamlit-based UI for interactive control
• —🔁 Iterative feedback cycles supported at all stages

⚙️ Architecture

The system is built using LangGraph's stateful graph model. Each phase of SDLC is a graph node, powered by a corresponding LLM agent. Here's how it works:

User Input ➝ User Story Generator ➝ Document Creator ➝ Document Writer ➝ Document Reviewer ➝ Code Generator ➝ Reviewer ➝ Security ➝ Test Writer ➝ Final Output

Tech Stack

• —🧩 LangGraph – to model the SDLC workflow as a graph
• —🧠 LLMs – OpenAI (gpt-4o, gpt-4) or Groq models (llama-3, gemma)
• —🎛️ Streamlit – for user feedback, config, and step-wise UI
• —🗂️ Python – Modular, extensible structure

🛠️ Installation

1. 1.Clone the repo:

git clone https://github.com/your-username/agentic-sdlc.git cd agentic-sdlc

1. 1.Install dependencies:

pip install -r requirements.txt

1. 1.Set up environment variables in a .env file:

OPENAIAPIKEY=youropenaikey GROQAPIKEY=yourgroqkey

1. 1.Run the app:

streamlit run app.py

📷 Mermaid Diagram

graph TD;
	__start__([<p>__start__</p>]):::first
	User_Story_Generator(User Story Generator)
	Product_Owner(Product Owner)
	Document_creator(Document creator)
	Document_writer(Document writer)
	Document_compiler(Document compiler)
	Document_Reviewer(Document Reviewer)
	Code_Generator(Code Generator)
	Code_Reviewer(Code Reviewer)
	Security_Reviewer(Security Reviewer)
	Test_Writer(Test Writer)
	Testcase_reviewer(Testcase reviewer)
	__end__([<p>__end__</p>]):::last
	Code_Generator --> Code_Reviewer;
	Document_compiler --> Document_Reviewer;
	Document_writer --> Document_compiler;
	Test_Writer --> Testcase_reviewer;
	User_Story_Generator --> Product_Owner;
	__start__ --> User_Story_Generator;
	Product_Owner -. &nbsp;Accepted&nbsp; .-> Document_creator;
	Product_Owner -. &nbsp;Feedback&nbsp; .-> User_Story_Generator;
	Document_creator -.-> Document_writer;
	Document_Reviewer -. &nbsp;Accepted&nbsp; .-> Code_Generator;
	Document_Reviewer -. &nbsp;Feedback&nbsp; .-> Document_creator;
	Code_Reviewer -. &nbsp;Accepted&nbsp; .-> Security_Reviewer;
	Code_Reviewer -. &nbsp;Reiterate&nbsp; .-> Code_Generator;
	Security_Reviewer -. &nbsp;Accepted&nbsp; .-> Test_Writer;
	Security_Reviewer -. &nbsp;Reiterate&nbsp; .-> Code_Generator;
	Testcase_reviewer -. &nbsp;Accepted&nbsp; .-> __end__;
	Testcase_reviewer -. &nbsp;Reiterate&nbsp; .-> Test_Writer;
	classDef default fill:#f2f0ff,line-height:1.2
	classDef first fill-opacity:0
	classDef last fill:#bfb6fc

✍️ Feedback Loop Support

At any stage (user stories, design, code, test), if the human reviewer provides feedback, the pipeline automatically routes back and regenerates output — just like a real review cycle.

📦 Folder Structure

src/ SDLC/ ui/ → Streamlit UI LLM/ → Model setup and orchestration graph/ → LangGraph builder nodes/ → Individual agent logic (story, doc, code, test) state/ → TypedDict + Pydantic-based state

🧩 Customization

You can plug in other models or modify the stages easily. Just update the LLMconfig and SDLC_Nodes to reflect new capabilities.

🙌 Credits

LangGraph

LangChain

Streamlit

📬 Contributions

PRs welcome! Please open an issue first to discuss major changes.