AGENTS.md
AGENTS.md is a structured project context file designed specifically for AI assistants.
It acts as a single source of truth that explains how your project works — not for humans, but for AI tools.
Instead of repeatedly describing your stack, architecture, conventions, and patterns in every prompt, you define them once inside AGENTS.md.
AI tools can then use this file as a persistent reference when generating code, refactoring features, or answering project-related questions.
Learn more about the concept at: https://agents.md/
Why NativeLaunch includes it
NativeLaunch is built as an AI-first starter.
Modern development increasingly relies on tools such as:
- ChatGPT
- JetBrains AI Assistant
- GitHub Copilot
- Claude
- Cursor
- VS Code AI extensions
By default, most AI tools only see:
- the current file
- selected code
- limited surrounding context
Without structured project guidance, this often results in:
- inconsistent folder placement
- duplicated logic
- incorrect architecture assumptions
- breaking existing conventions
AGENTS.md solves this by explicitly defining:
- project architecture
- feature boundaries
- state management approach
- styling conventions
- file and folder structure
- naming patterns
- integration patterns (auth, payments, analytics, notifications)
This leads to:
- higher quality code generation
- architectural consistency
- fewer refactors later
- faster feature development
How to use it with AI tools
When working inside your IDE:
- Open the AI chat panel
- Ask the assistant to read
AGENTS.md - Use prompts such as:
- "Use AGENTS.md as project context"
- "Follow the architecture defined in AGENTS.md"
- "Generate a feature aligned with AGENTS.md"
You can also reference it explicitly:
According to AGENTS.md, where should this feature live?
This ensures that generated code respects your existing structure and conventions.
When should you update it?
Update AGENTS.md whenever you:
- Introduce a new major feature
- Change architecture or folder structure
- Add new integrations
- Refactor core systems
- Change coding standards
Think of it as:
README for AI contributors.
Why this matters
AI-assisted development is no longer experimental — it is part of daily workflows.
Projects that provide structured AI context:
- Scale faster
- Maintain consistency
- Reduce onboarding time
- Lower cognitive load
AGENTS.md turns your project into a system that both humans and AI can understand.
And that is a long-term advantage.