GoBetter Documentation
GoBetter is an autonomous code review engine and developer intelligence platform designed to deliver high-signal, deep architectural pull request reviews. It intercepts GitHub PR webhooks, isolates diff context, coordinates background review workers across distributed queues, and posts actionable findings directly to your pull requests.Explore GoBetter
Supercharge Your AI Coding Agents with GoBetter MCP
Give your AI coding assistants instant access to GoBetter docs, architecture specs, and API schemas—right inside your editor.
Add manually & CLI (Cursor, VS Code, Claude Code, Codex)
Add manually & CLI (Cursor, VS Code, Claude Code, Codex)
- Cursor
- VS Code
- Claude Code
- Codex CLI
Add the GoBetter MCP server to your Cursor
mcp.json (Cmd/Ctrl + Shift + P → Open MCP settings):Quickstart
Set up and run GoBetter locally with PostgreSQL, Redis, and Vite in under 5 minutes.
Architecture
Explore the webhook ingestion lifecycle, BullMQ queues, and state handler transitions.
Server Configuration
Configure backend environment variables, database connections, and AI provider routing.
Client Configuration
Customize frontend runtime settings, local storage toggles, and feature flags.
Core Capabilities
High-Signal Review Engine
High-Signal Review Engine
GoBetter is tuned for zero-nitpick reviews. Rather than spamming pull requests with style or formatting comments, the agent specifically targets critical architectural flaws, security risks, concurrency defects, and logic regressions.
Bring Your Own Key (BYOK) & Custom Models
Bring Your Own Key (BYOK) & Custom Models
Connect your own API keys for OpenAI, Anthropic Claude, Google Gemini, DeepSeek, or custom OpenAI-compatible enterprise gateways. All API keys are encrypted at rest using symmetric AES-256-GCM.
Distributed Async Architecture
Distributed Async Architecture
Webhook payloads are immediately acknowledged and dispatched to Redis-backed BullMQ queues. Independent workers handle payload sanitization, resource extraction, AI inference, and notification dispatch.
Interactive Web Dashboard
Interactive Web Dashboard
Inspect pull request review histories, toggle automated reviews, configure prompt instructions per review mode (Quick, Standard, Deep Dive), and manage active sessions across devices.
Next Steps
Getting Started
Run the platform locally.
AI Harness
Learn how agents and skills are organized.
API Reference
Explore backend endpoints and schemas.