AI-Driven Development Framework

Your AI assistant, now a full SDLC expert

Smart SDLC gives GitHub Copilot, Claude, Cursor, and any AI assistant structured expertise across analysis, architecture, stories, implementation, and modernization — no runtime required.

30+
Skills
6
AI Personas
5
Integrations
Zero
Dependencies
MIT
License
Smart SDLC AI Framework
Open Source: Smart SDLC is MIT licensed and free to use. A standalone AI-driven agile development framework that works natively inside GitHub Copilot, Claude, or any AI coding assistant. No external runtime — your AI assistant IS the engine.

What It Is

Structured AI Expertise Across Your Entire SDLC

Each skill is a markdown file (SKILL.md) containing structured instructions the AI reads and follows. No code execution, no installed tooling — just skills your AI activates on demand.

Skills Framework

Markdown-based skill files that give your AI assistant deep, structured knowledge for each SDLC phase.

Works With Any AI

Native support for GitHub Copilot, Claude, Cursor, and any AI assistant that can read files.

Zero Dependencies

Pure Node.js CLI — no external packages required. Install and run instantly with npx.

Team-Ready

Project-wide setup once by the team lead, then each member configures their own persona independently.

Capabilities

Everything Your Team Needs to Ship Faster

Smart SDLC covers the entire software development lifecycle with native integrations to your existing tools.

Persona-Based Agents

5 AI personas (Product, Architect, Developer, Modernization Lead, Team Lead) — each with a distinct role and skill set. Appear natively in the GitHub Copilot agent picker.

Full SDLC Coverage

Skills for analysis, requirements (PRD), architecture, ADRs, epics & stories, implementation, sprint planning, and modernization.

Native Integrations

JIRA, Confluence, GitHub, GitLab, and Azure DevOps workflows — via REST API or MCP Server.

Company Knowledge

Register internal framework docs, platform libraries, and developer portals — fetched and pulled into AI context on demand.

Conflict Prevention

JIRA ticket lock, git branch lock, and Confluence version traceability prevent parallel work conflicts across team members.

Multi-Persona Meetings

Generate a structured context prompt bringing all personas into one AI session for design reviews and sprint planning.

Copilot Slash Commands

Skills are generated as .github/skills//SKILL.md — each appears as a /skill-name slash command in GitHub Copilot Chat.

Custom Role Names

Your team names each persona (e.g. "Aria", "Rex", "Nova") — names flow through meeting docs and agent greetings.

AI Personas

Meet Your AI Development Team

Six specialized AI personas, each with deep expertise in their role. Fully customizable names and communication styles.

Product / BA — Aria

Requirements, PRDs, user stories. Activate via @sml-agent-pm in GitHub Copilot.

Architect — Rex

System design, ADRs, architecture decisions. Activate via @sml-agent-architect.

Developer — Nova

Implementation, code review, tech debt tracking. Activate via @sml-agent-developer.

Modernization Lead — Sage

Legacy analysis, migration planning. Activate via @sml-agent-sage.

Team Lead / PM — Lead

Epics, sprint planning, delivery coordination. Activate via @sml-agent-lead.

Code Archaeologist — Scout

Codebase onboarding, reverse-engineer docs. Activate via @sml-agent-scout.

Get Started in Minutes

Set up Smart SDLC for your entire team with a few simple commands.

Step 1: Set up the project

Run once — your team lead or PM:

npx @supermldev/smart-sdlc init
Installs skills, creates config, scaffolds reference folders, and generates Copilot agent/skill files.

Step 2: Set up your personal workspace

Each team member runs once:

npx @supermldev/smart-sdlc persona
Configures your name, role, AI tool preference, and skill level.

Step 3: Add company reference docs

Drop markdown files into _superml/reference/ — shared docs go in all/, role-specific docs in the role subfolder. Agents load these automatically on every activation.

Step 4: Activate an agent in your AI chat

In GitHub Copilot Chat:

@sml-agent-pm
In Claude, Cursor, or any AI assistant:

Load the skill at _superml/skills/2-planning/agent-pm/SKILL.md

Step 5: Run a multi-persona meeting

Bring your whole AI team together:

npx @supermldev/smart-sdlc meeting
Generates a structured context prompt for design reviews and sprint planning.

Team working together

CLI Reference

All Commands at a Glance

Smart SDLC's CLI covers every phase of team setup and ongoing usage.

$ npx @supermldev/smart-sdlc init       # Set up Smart SDLC project for your team
$ npx @supermldev/smart-sdlc persona    # Configure your personal workspace and role
$ npx @supermldev/smart-sdlc help       # Context-aware SDLC guidance
$ npx @supermldev/smart-sdlc list       # List all available skills and agents
$ npx @supermldev/smart-sdlc meeting    # Set up a multi-persona meeting context
$ npx @supermldev/smart-sdlc update     # Update skills to the latest version
$ npx @supermldev/smart-sdlc clean      # Remove generated Smart SDLC files
      

Integrations

JIRA, Confluence, GitHub, GitLab, Azure DevOps — via REST API or MCP Server.

Conflict Prevention

JIRA ticket lock, git branch lock, and Confluence version traceability built in.

Always Up to Date

Run update anytime to pull the latest skills without re-running init.

FAQ

Frequently Asked Questions

Everything you need to know about Smart SDLC.

Does Smart SDLC require any external runtime or services?

No. Smart SDLC is pure Node.js with zero external dependencies. Skills are plain markdown files — your AI assistant reads them directly. No servers, no APIs, no subscriptions required.

Which AI assistants are supported?

Smart SDLC works with GitHub Copilot (with native agent picker support), Claude, Cursor, and any AI assistant that can read files or accept pasted context.

Can I customize the persona names and roles?

Yes. During init you name each persona for your team (e.g. "Aria", "Rex"). Those names flow through all generated docs, meeting prompts, and agent greetings automatically.

How do company-specific docs get into the AI context?

Drop markdown files into _superml/reference/ — in all/ for shared docs or a role subfolder for role-specific content. Agents load them automatically every time they are activated.

How does conflict prevention work?

Three layers: JIRA ticket lock (checks In Progress + assigned status before starting a story), git branch lock (checks remote branch name before creating), and Confluence version traceability (embeds source doc version in stories so stale-doc conflicts surface early).

Is Smart SDLC free to use?

Yes. Smart SDLC is open source under the MIT license — free for personal and commercial use.

Community

Join the Smart SDLC Community

Connect with other teams using AI-driven development, share workflows, and contribute new skills.

Discord

Ask questions, share workflows, and connect with other Smart SDLC users and contributors on Discord.

GitHub

Browse the source, open issues, and contribute new skills and integrations on GitHub.

Contributing

New skills, integrations, CLI commands, and documentation improvements are all welcome. See CONTRIBUTING.md for guidelines.

Ready to give your AI assistant full SDLC expertise?

Set up Smart SDLC in minutes with a single npx command. Free, open source, zero dependencies.