claude-lrc

Free, Micro AI code reviews that run from Claude Code

 

---

AI agents write code fast. They also silently remove logic, change behavior, and introduce bugs -- without telling you. You often find out in production.

claude-lrc fixes this. It brings AI code review directly into Claude Code and reviews every diff before it lands. 60-second setup. Completely free.

Review, vouch, or skip — without switching tools or remembering separate CLI steps. Just tell Claude what you want, or use a slash command. Either way, the review runs.

See It In Action

claude-lrc1-final.mp4

Why

• 🤖 AI agents write code fast and quietly. Logic removed. Behavior changed. Edge cases gone. You notice in production.
• 🔍 Review while the change is fresh. AI inline comments show you exactly what changed and what looks wrong — before it lands.
• 📍 Review where the code was generated. You're already in Claude Code. The review stays there too.
• 🗣️ Natural language lowers friction. No separate CLI to remember. Just say what you want.
• 🎯 Explicit commands give deterministic control. When you want exact behavior, slash commands deliver it.
• 🔁 Teams and individuals both benefit. The review habit is built in, not dependent on everyone remembering a step.

Prerequisites

claude-lrc uses the LiveReview infrastructure for AI code review — the same backend that powers git-lrc.

You don't need git-lrc pre-installed. The plugin handles its own backend bootstrap during setup. But if you already use git-lrc, claude-lrc fits right alongside it — they share the same review state, git log entries, and coverage tracking.

Also Included: Git Commit Workflow

🔗 We also bundle git-lrc alongside this workflow. Use claude-lrc when you want review, vouch, skip, and hook control from natural language or slash commands inside Claude Code. Use git-lrc when you want git-native, editor-independent commit-time enforcement across repos and tools.

Get Started

Install

Add the marketplace:

claude plugin marketplace add HexmosTech/claude-lrc

Install the plugin:

claude plugin install lrc@claude-lrc

Reload plugins in the current Claude session:

/reload-plugins

Setup

Run setup directly:

/lrc:setup

Or just ask Claude:

set up lrc

Setup opens browser-based steps for:

1. LiveReview account — sign in with Hexmos
2. AI connector — connect a Gemini, OpenAI, Claude, or other supported API key

One-time setup. After this, review flows are ready in every repo you open in Claude Code.

How You Use It

Option A: Natural language

Tell Claude what you want:

review with lrc
skip lrc review
vouch for this
check if lrc is active
enable lrc hooks
disable claude gate

The plugin routes those requests to the matching review or hook flow.

Option B: Slash commands

Use explicit slash commands when you want exact, predictable behavior:

/lrc:review
/lrc:skip
/lrc:vouch
/lrc:hooks
/lrc:setup

What You Can Do

Review code changes

Stage your files, then ask for a review:

review with lrc

Or explicitly:

/lrc:review

The review UI opens in your browser. AI comments appear inline on the diff.

Skip review

Intentionally bypass AI review and record that decision:

skip lrc review

Or:

/lrc:skip

The git log will record skipped — no review ran, and no attestation is made.

Vouch manually

You've iterated, you're satisfied, and you want to take responsibility without triggering another AI pass:

vouch for this

Or:

/lrc:vouch

Coverage and iteration stats from prior review rounds are preserved.

Manage hook state

check if lrc is active
enable lrc hooks
disable claude gate

Or use the explicit hook command:

/lrc:hooks

The Review Experience

When a review runs, the UI opens in your browser. You get:

• GitHub-style diff — color-coded additions and deletions
• Inline AI comments — at the exact lines that matter, with severity badges
• Review summary — high-level overview of what the AI found
• Staged file list — all staged files, with navigation between them
• Copy issues — one click to copy all AI-flagged issues, ready to paste back into Claude
• Cycle through issues — navigate between comments without scrolling

The Review Cycle

Typical workflow with AI-generated code:

1. Ask Claude to generate or modify code.
2. Stage the files you want reviewed.
3. Tell Claude review with lrc.
4. Inspect the browser review — inline AI comments, summary, file list.
5. Copy the flagged issues and feed them back to Claude to fix.
6. Stage the fixes and run review with lrc again.
7. Repeat until satisfied.
8. Use vouch for this or let Claude commit through the reviewed flow.

Each review with lrc is one iteration. The tool tracks how many iterations you did and what percentage of the final diff was AI-reviewed (coverage).

Git Log Tracking

Every commit gets a review status line in the git log:

LiveReview Pre-Commit Check: ran (iter:3, coverage:85%)

LiveReview Pre-Commit Check: vouched (iter:2, coverage:50%)

LiveReview Pre-Commit Check: skipped

• iter — number of review cycles before committing
• coverage — percentage of the final diff that was AI-reviewed in prior iterations

This is shared with git-lrc. If you use both, you'll see a consistent review history across your git log regardless of which tool ran the review.

Review vs Vouch vs Skip

	Review	Vouch	Skip
AI reviews the diff?	✅ Yes	❌ No	❌ No
Takes responsibility?	✅ Yes	✅ Yes, explicitly	⚠️ No
Tracks iterations?	✅ Yes	✅ Records prior coverage	❌ No
Git log message	ran (iter:N, coverage:X%)	vouched (iter:N, coverage:X%)	skipped
When to use	Each review cycle	Done iterating, ready to commit	Not reviewing this commit

Review is the default. AI analyzes your staged diff and gives inline feedback.

Vouch means you're explicitly taking responsibility for this commit. Typically used after multiple review iterations — you've gone back and forth with Claude, fixed issues, and are now satisfied. No AI review runs, but your prior iteration and coverage stats are preserved.

Skip means you're not reviewing this particular commit. The git log records skipped.

Quick Reference

Input	What happens
review with lrc or review this	Review staged changes
/lrc:review	Review staged changes (explicit)
skip lrc review	Skip AI review, write attestation
/lrc:skip	Skip review (explicit)
vouch for this	Manual vouch without AI review
/lrc:vouch	Manual vouch (explicit)
check if lrc is active	Show hook status
enable lrc hooks	Enable hooks
disable claude gate	Disable Claude-surface hooks
/lrc:hooks	Hook management entry point
/lrc:setup	Interactive setup flow

Bring Your Own AI Connector (BYOK)

By default, setup connects Gemini (Google AI Studio). You can also connect:

• OpenAI
• Claude
• DeepSeek
• OpenRouter

To manage connectors, run:

lrc ui

From the UI, you can:

• Re-authenticate your account
• Add or update AI connectors
• Reorder connectors to set priority

The first connector in the list is used for reviews.

Community

• Discord: discord.gg/sGdnKwB3qq — quick questions, general Q&A, back-and-forth with the team
• GitHub Issues: github.com/HexmosTech/claude-lrc/issues — bugs and scoped feature requests
• git-lrc repo: github.com/HexmosTech/git-lrc — the CLI counterpart and shared backend

License

claude-lrc is distributed under a modified variant of the Sustainable Use License (SUL).

Note

What this means:

• ✅ Source Available — Full source code is available for self-hosting
• ✅ Business Use Allowed — Use claude-lrc for your internal business operations
• ✅ Modifications Allowed — Customize for your own use
• ❌ No Resale — Cannot be resold or offered as a competing service
• ❌ No Redistribution — Cannot redistribute modified versions commercially

This license ensures the project remains sustainable while giving you full access to self-host and customize for your needs.

For detailed terms, examples of permitted and prohibited uses, and definitions, see the full LICENSE.md.

---

For Teams: LiveReview

Using claude-lrc solo? Great. Building with a team? Check out LiveReview — the full suite for team-wide AI code review, with dashboards, org-level policies, and review analytics. Everything claude-lrc does, plus team coordination.