Setup guide: Cursor + Claude Code + Copilot

You can set up RepoWise with Cursor, Claude Code, and GitHub Copilot in under 5 minutes. Here is how.

Step 1: Install and run

Open your terminal in the project directory and run:

RepoWise prompts you to authenticate with GitHub, GitLab, or Bitbucket. Select your repository. The CLI handles the rest.

No global installs. No config files to write. One command.

Step 2: The AI interview (optional)

After selecting your repo, RepoWise asks if you want to do a quick interview. This is optional but makes a measurable difference in context quality.

The interview covers three areas:

• Architecture decisions. Why you chose your tech stack, what patterns you follow, which modules own which responsibilities.
• Business rules. Domain-specific logic that code alone does not explain. Pricing calculations, compliance requirements, edge cases.
• Team conventions. Naming patterns, code organization rules, testing expectations, PR review standards.

The interview takes 2-3 minutes. You can skip it and run it later from the dashboard.

Step 3: Context generated

RepoWise analyzes your codebase and generates structured context files. The first scan takes a few minutes depending on repo size.

When it finishes, your context files download to:

Inside that directory you will find structured files covering architecture, data models, API contracts, coding patterns, domain knowledge, security practices, and more. Each file focuses on one topic. No giant blob.

Every AI tool picks it up automatically

Context downloads to your project directory. Cursor, Claude Code, Copilot, and every other supported AI tool reads the context files natively. No manual copying. No format conversion. No configuration.

Automatic sync

After the first scan, RepoWise keeps your context fresh automatically. Every merge triggers an incremental update. You never touch it again. Every AI tool on your team reads the updated context immediately.

Using multiple tools

Switch between Cursor and Claude Code freely. The context is the same because it comes from the same source of truth. Add Copilot later and it picks up the same context instantly.

New team members clone the repo, run one command, and every AI tool they use has full context from the start. No onboarding docs to read. No senior engineer to shadow for a week.

One scan, every tool, always fresh

That is the full setup. One command to generate. Automatic sync on every merge. Every AI tool on your team reads the same accurate context about your codebase. Try it: