[Resource]: Multi-Agent Code Review: Open Code Review

[spencermarx]

Display Name

Open Code Review

Category

Tooling

Sub-Category

Tooling: Orchestrators

Primary Link

https://github.com/spencermarx/open-code-review

Author Name

Spencer Marx

Author Link

https://github.com/spencermarx

License

Apache-2.0

Other License

No response

Description

TLDR: Multi-agent code review modeled after high performing engineering teams.

Open Code Review is a Claude Code or Agentic IDE tool that allows one to configure a team of AI code reviewers (via sub-agents). Architecture, code quality, security, testing, or custom roles that fit your codebase. They review independently and then debate each other before you ever see the final code review report.

Validate Claims

CLI Installation

Works with any AI coding assistant (Claude Code, Cursor, Windsurf, GitHub Copilot, and more). Provides progress tracking and multi-tool configuration.

# Install the CLI
npm install -g @open-code-review/cli

# Initialize in your project
cd your-project
ocr init

The CLI detects your installed AI tools and configures each appropriately.

Running Your First Code Review

1. Stage your changes and start a review:

git add .

Then in Claude Code:

/ocr:review compare the staged changes against my requirements described in `spec.md`

2. (Optional) Watch progress in real-time:

In a separate terminal using the CLI:

ocr progress

  Open Code Review

  2026-01-26-main  ·  Round 2  ·  1m 23s

  ━━━━━━━━━━━━━━━━────────  60%  ·  Parallel Reviews

  ✓ Context Discovery
  ✓ Change Context
  ✓ Tech Lead Analysis
  ▸ Parallel Reviews
    Round 2
    ✓ Principal #1 2  │  ✓ Principal #2 1  │  ○ Quality #1 0  │  ○ Quality #2 0
    Round 1 ✓ 4 reviewers
  · Aggregate Findings
  · Reviewer Discourse
  · Final Synthesis

  Ctrl+C to exit

3. Review the output:

# Code Review: Feature/Auth Implementation

## Verdict: ✅ APPROVE with suggestions

### Critical (0)
No blocking issues.

### Suggestions (3)
1. **Add rate limiting** — Auth endpoints lack rate limiting
2. **Token expiry** — Consider shorter JWT expiry for security
3. **Error messages** — Avoid leaking user existence in login errors

### Requirements Verification
| Requirement | Status |
|-------------|--------|
| JWT authentication | ✅ Implemented |
| Refresh tokens | ✅ Implemented |
| Password hashing | ✅ Using bcrypt |

Specific Task(s)

Have it perform a code review of a change to this repo

Specific Prompt(s)

The key prompt is:

/ocr:review our staged changes against the requirements described in `spec.md` (or provide inline requirements description)

Additional Comments

• You can customize your code review context in .ocr/config-yaml
• Customizations include:
  • Where to source key project context from (beyond the defaults)
  • Custom reviewers (for sub-agent spawning) that can be used by the tech lead -- in .ocr/skills/references/reviewers
  • Reviewer redundancy -- configure multiple sub-agents for a given reviewer type
• Multi-round review architecture (following the standard iterative code review process we all know and love)

Recommendation Checklist

• I have checked that this resource hasn't already been submitted
• My resource provides genuine value to Claude Code users, and any risks are clearly stated
• All provided links are working and publicly accessible
• I am submitting only ONE resource in this issue
• I understand that low-quality or duplicate submissions may be rejected

[github-actions]

🤖 Validation Results

✅ All validation checks passed!

Your submission is ready for review by a maintainer.

Validated Data:

{
  "display_name": "Open Code Review",
  "category": "Tooling",
  "subcategory": "Orchestrators",
  "primary_link": "https://github.com/spencermarx/open-code-review",
  "author_name": "Spencer Marx",
  "author_link": "https://github.com/spencermarx",
  "license": "Apache-2.0",
  "description": "TLDR: Multi-agent code review modeled after high performing engineering teams.\nOpen Code Review is a Claude Code or Agentic IDE tool that allows one to configure a team of AI code reviewers (via sub-agents). Architecture, code quality, security, testing, or custom roles that fit your codebase. They review independently and then debate each other before you ever see the final code review report.",
  "secondary_link": "",
  "last_modified": "2026-01-28:11-18-36",
  "active": "TRUE",
  "last_checked": "2026-01-28:12-09-34"
}

---

_{This comment is automatically updated when you edit the issue.}

[hesreallyhim]

thanks @spencermarx ! appreciate the recommendation. just letting you know ahead of time we do have a few multi-agent review things already, probably because it's a very good thing to do.

[spencermarx]

Thanks for the note @hesreallyhim and I completely agree 🤘

From my view, the future of AI-assisted dev has multiple agent-based code review as a key part of the workflow.

And hey! Really appreciate you taking a look at this project 🙂

When I sat down to design it, I really wanted it to stay out of the way of folks and their existing workflows. Hopefully that was achieved (friends/colleagues using this seem to think so so far).

It essentially boils down to an orchestrator sequence with parameters for context (project + code review, etc.), reviewer types, and reviewer redundancy. No opinions are forced, mainly a structured protocol for how to build context, spawn agents, and distill their response back into a useful clear feedback document.

[spencermarx]

Happy Friday @hesreallyhim 🙂

Just a note, yesterday we pushed up an additional feature that could be of interest:

New command: /ocr:map

Code Review Map generation for human reviewers. Due diligence and code review tracking document creation performed by an architect orchestrator + multi-agent codebase researchers. Meant to help a human tackle a large change set sanely and in an organized way.

Already using it on our team to help us work through some of the larger PRs that sit in our PR queue (another side effect of AI-assisted dev).