io.github.selvage-lab/selvage

About this integration

An LLM-based code review MCP server with AST-powered smart context extraction

Transport

stdio

Authentication

unknown

Initial setup

unknown

Runtime

unattended

Evidence

documented

Version

0.4.1

Package

selvage

Last compatibility test

Not independently tested

Connect your agent

Publisher documentation reviewed; package, Git operations, and model providers were not executed. Authentication is mode-dependent: delegated get_review_context uses the host agent and needs no Selvage API key, while provider-backed review modes use OpenRouter or individual provider API keys. The native authentication taxonomy does not express this mixed optional-credential surface, so authentication and setup remain unknown rather than implying every tool is unauthenticated or credentialed.

Capabilities: Prepare Git diffs and AST-based context for code review, Delegate review to the host agent without a Selvage API key, Run provider-backed multi-model code reviews when credentials are configured, Review staged, unstaged, branch, or commit-based changes

Connected profiles

Additional details