Skip to main content
BugViper connects to your GitHub repositories, parses every function, class, and file into a searchable knowledge graph, then deploys an AI agent that reviews your pull requests with full codebase context — posting inline comments on the exact lines where issues were found.

Quick Start

Index your first repository and trigger an AI review in minutes.

Set Up GitHub App

Install the BugViper GitHub App to enable webhook-driven reviews.

Code Search

Search your entire codebase by symbol name, content, or natural language.

API Reference

Integrate BugViper into your workflow with the REST API.

​
What BugViper does

When you add a repository, BugViper parses it into a Neo4j knowledge graph — every function, class, variable, file, and import becomes a node with relationships like CALLS, IMPORTS, and INHERITS. This graph powers three core capabilities:

PR Reviews

Mention @bugviper review on any pull request to get inline bug and security comments posted automatically.

Code Search

Full-text and semantic search across your entire indexed codebase, with a line-level peek viewer.

Ask Agent

Ask natural language questions about your codebase and get answers with cited source code.

​
Get started in four steps

1

Install the GitHub App

Install BugViper on your GitHub account or organization and grant access to your repositories.
2

Configure environment

Set your OpenRouter API key, Neo4j connection, and GitHub App credentials.
3

Index a repository

Use the dashboard or API to ingest a repository. BugViper clones it, parses it with Tree-sitter, and writes the graph.
4

Trigger your first review

Open a pull request and comment @bugviper review. The AI agent will post inline findings within minutes.
BugViper supports 17 programming languages including Python, TypeScript, JavaScript, Go, Rust, Java, and C++. See the full language list.