Skip to main content
BugViper is a GitHub App that brings codebase-aware AI code review to every pull request. Once installed, BugViper clones your repo into a secure E2B sandbox, builds a live call graph with tree-sitter to map blast radius across files, then runs DeepAgents — or fast static analysis — to post inline comments, track issues, and surface analytics, all without touching your local environment.

Quick Start

Install the GitHub App and get your first AI review in under 5 minutes.

Bot Commands

All commands you can trigger by commenting on a pull request.

Dashboard

Browse repos, inspect reviews, and track analytics across your team.

API Reference

Explore the REST API for webhooks, analytics, and tool configuration.

​
Get up and running

BugViper is designed to go from zero to first review in a single sitting. Follow these four steps and you’ll have AI-powered inline comments on your next pull request.
1

Install the GitHub App

Go to the GitHub Marketplace or your BugViper dashboard and install the BugViper GitHub App on your organization or personal account. Select which repositories BugViper can access and approve the required permissions.
2

Sign in with GitHub

Open the BugViper dashboard and sign in using GitHub OAuth. Your account is linked to the GitHub App installation you just authorized.
3

Open a PR and comment @bugviper

On any pull request in a connected repository, leave a comment with @bugviper full review to trigger a full AI review, or @bugviper run lint for a fast static-analysis pass.
4

Review inline comments and analytics

BugViper posts inline comments on the relevant lines, adds a summary review body, and logs everything to your dashboard — bugs caught, issues tracked, and merge-time analytics updated automatically.

​
Features

BugViper ships with a full suite of review and tracking capabilities out of the box.

Call Graph Analysis

Builds a live call graph via tree-sitter so agents see cross-file function interactions and real blast radius, not just the isolated diff.

Deep Review Mode

Three specialized sub-agents — Bug, Security, and Performance — run in parallel inside isolated E2B sandboxes for thorough, multi-angle analysis.

Static Analysis

Runs ESLint, Ruff, and golangci-lint against changed files for a fast lint-only pass covering Python, JavaScript/TypeScript, Go, and more.

False Positive Filter

A dedicated verifier pass validates every finding against the actual diff and classifies issues as valid, nitpick, or outside_diff before posting.

Issue Tracking

Full issue lifecycle management — open, resolved, and fixed — with per-issue GitHub comment linkage and automatic resolve detection across PRs.

Analytics Dashboard

Interactive charts for bugs caught, addressed rate, average merge time, and PRs reviewed per day, broken down per repository.