rvue extracts coding conventions, anti-patterns, and domain knowledge from your PR history and delivers it to every AI coding tool automatically.
Free for open source. No credit card required.
Delivers intelligence to your AI coding tools
How it works
See exactly what rvue extracts from your codebase and where it delivers it.
One command reads your codebase, linter configs, and merged PRs with every review comment.
AI identifies the conventions your team enforces, anti-patterns that get code rejected, and domain knowledge that only lives in people's heads.
An Agent Skill, MCP server, and AGENTS.md that works with Cursor, Claude Code, Copilot, and any MCP-compatible tool.
See the difference
Same AI tool, same request. The only difference is your team's intelligence.
export default async function handler(
req, res
) {
const { email, name } = req.body
const user = await db.user.create({
data: { email, name },
})
console.log("Created:", user.id)
return res.status(200).json({ user })
}export const createUser = async (
req: Request
) => {
const input = CreateUserSchema.parse(
await req.json()
)
const user = await db.user.create({
data: input,
})
return Response.json(
{ user }, { status: 201 }
)
}Built for teams
Built on open standards, privacy-first, and designed to stay out of your way.
Agent Skills are auto-discovered. Teammates clone and code. Intelligence stays in your repo, always up to date.
Analysis happens on our servers but no source code is stored. Only extracted patterns and conventions are persisted.
Built on the open Agent Skills standard plus MCP protocol. Native support for Cursor, Claude Code, Copilot, and any MCP client.
A GitHub Action keeps intelligence fresh. Every merged PR refines your team's conventions automatically.
Every convention is scored by evidence: how many PRs enforce it and how consistently it appears across the codebase.
Built on agentskills.io. Your intelligence is portable, version-controlled, and not locked into any single tool.