← Back to Downloads

Getting Started with Scrubby

From installation to your first code review in minutes

1 Installation

Download the .vsix file from the staging release (requires GitHub repo access).

Via VS Code UI:

  1. Open Extensions (Cmd+Shift+X / Ctrl+Shift+X)
  2. Click the ... menu and select Install from VSIX...
  3. Pick the downloaded file, then reload VS Code when prompted

Via CLI:

code --install-extension scrubbyai-0.6.0-staging.vsix

2 Sign In with GitHub

Scrubby uses GitHub OAuth for authentication — no API keys needed.

  1. Open the Command Palette (Cmd+Shift+P / Ctrl+Shift+P)
  2. Run ScrubbyAI: Sign In with GitHub
  3. Authorize in the browser pop-up
  4. You'll see a confirmation in the status bar once signed in
Your GitHub token is exchanged for a Scrubby session behind the scenes. AI features are proxied through the Scrubby backend — you never need to manage an Anthropic API key.

3 Setup Wizard

On first launch (or via ScrubbyAI: Run Setup Wizard), a guided wizard walks you through:

You can re-run the wizard or change these settings at any time.

4 Using Scrubby

Real-time linting — Once enabled, Scrubby highlights issues inline as you type. Hover over underlined code for explanations and suggested fixes.

AI chat — Open the Scrubby sidebar panel to ask questions about your code, request refactors, or get explanations. The chat is context-aware and understands your workspace.

Pattern violations — Scrubby detects when new code deviates from your repository's established patterns and suggests how to align it.

Static analysis — Run a full project scan with ScrubbyAI: Run Static Analysis for a comprehensive report across all files.

5 Commands

All commands are available from the Command Palette (Cmd+Shift+P):

Run Setup Wizard Guided first-run configuration
Sign In with GitHub Authenticate via GitHub OAuth
Enable Real-time Linting Turn on inline analysis
Disable Real-time Linting Turn off inline analysis
Discover Code Patterns AI scans for repo conventions
Check Pattern Violations Find code deviating from patterns
Assess Pattern Quality Rate discovered pattern confidence
Run Static Analysis Full project scan
Run SME Analysis Framework/concern expert analysis
Install Git Hooks Add pre-commit pattern checks
Uninstall Git Hooks Remove pre-commit hooks
Open ScrubbyAI Panel Open the AI chat sidebar
Edit Configuration Open or create scrubby.yml
Check for Updates Check for a newer extension version
Account Status View your auth and plan info
All commands are prefixed with ScrubbyAI: in the Command Palette.

6 Configuration

Open VS Code Settings (Cmd+,) and search for ScrubbyAI:

ScrubbyAI.language Primary language: javascript, python, or ruby
ScrubbyAI.tone Suggestion style: friendly, professional, or concise
ScrubbyAI.enabled Toggle real-time linting on/off
ScrubbyAI.autoFix Auto-apply simple fixes (default: off)
ScrubbyAI.excludePaths Glob patterns to exclude (node_modules, dist, etc.)
ScrubbyAI.excludeFiles Specific files to skip (e.g. *.min.js)
ScrubbyAI.aiModel Claude model for AI chat (Sonnet 4.5 default)
You can also configure per-project settings in a scrubby.yml file at your workspace root. Run ScrubbyAI: Edit Configuration to create or open it.

7 Git Integration

Pre-commit hooks — Scrubby can install a Git pre-commit hook that checks staged files for pattern violations before each commit.

  1. Run ScrubbyAI: Install Git Hooks from the Command Palette
  2. Scrubby adds a hook to .git/hooks/pre-commit
  3. On each git commit, staged files are checked against your discovered patterns
  4. The commit is blocked if violations are found, with a summary of issues

To remove: run ScrubbyAI: Uninstall Git Hooks.

GitHub Actions — Run ScrubbyAI: Setup GitHub Actions Workflow to generate a CI workflow that analyzes PRs automatically.