2026 04 22 Claude Ultra Review

New in Claude Code: /ultrareview (research preview) runs a fleet of bug-hunting agents in the cloud.

Findings land in the CLI or Desktop automatically. Run it before merging critical changes—auth, data migrations, etc.

Pro and Max users get 3 free reviews through 5/5.

---

Run `claude update` to try it out. Learn more in the docs:

code.claude.com Find bugs with ultrareview - Claude Code Docs

---

Comments

PMtheBuilder @PMThebuilder · 2026-04-22

A fleet of bug-hunting agents running in parallel before you merge — this changes how a PM ships.

Before: you write the code, request a review, wait. Now: you build, run /ultrareview, a swarm stress-tests your auth flow and data migration in minutes while you move to the next

Jatin Garg @jatingargiitk · 2026-04-22

the interesting detail is "run it before merging critical changes." that's anthropic saying the bar for code review just moved. if your pr touches auth or migrations and you didn't run a fleet against it, the reviewer on your team will ask why. cultural expectations shift faster

Ray Browser @TheRayBrowser ·

Build it. Automate it. As much as you want. No cap. For a limited time.

raybrowser.com Unlimited credits. For a limited time.

Chen Avnery @MindTheGapMTG · 2026-04-22

We've been running multi-agent code review for months. Single-pass review catches syntax errors. Fleets catch the interactions between auth changes and data flows that actually break production. Glad to see this pattern going mainstream.

24AIGlobal @24AInor · 2026-04-22

running a fleet of agents to hunt bugs before merge is the right call for auth and migration code. still curious how it handles edge cases vs just static analysis patterns

Slavul Remist @thepatriotvlls · 2026-04-22

i used this yesterday and its the worst feature in existence. 1] it burns through your usage 2] QoR is abysmal since you cant go back and forth 3] its incredibly buggy and constantly disconnected from my terminal instance.

The Beyonder @Ultrabeyonder1 · 2026-04-22

Never in my life have i ever witness a company so clouded from reality… seriously get down from your stupid imaginary cloud of patheticness and watch whats REALLY going on with your clients… seriously if you dont fix or start doing something I will cancel my sub forever…

Tahseen Rahman @Tahseen_Rahman · 2026-04-22

This is the right direction. Once code agents can write large changes quickly, review has to become parallelized and adversarial enough to keep up.

surreal intelligence @Surreal_Intel · 2026-04-22

The intriguing possibility is that software teams end up with a permanent machine review layer sitting between “done” and “deployed”, quietly turning code shipping into a negotiation with synthetic auditors

Daniel Minda @mndaniel78 · 2026-04-22

Hi — if you want a read-only second pair of eyes on Claude Code output,

I built Witness — /review, no write/run access.

Uses your Claude login.

github.com GitHub - riskreadyeu/witness: Read-only AI code reviewer. Diff in, structured findings out. No...

Jon Mac @jonmacofficial · 2026-04-22

running this in our codebase now. it catches edge cases that grep-based review would miss. the cloud fleet model is the right architecture for distributed bug hunting.

Boakye Junior @junijas · 2026-04-22

Your creativity and innovation is undoubted. Most of your users are concerned about your subscription model and how tokens and limits run out so fast.

Diego Gutiérrez @guti10_FT · 2026-04-22

This is what it actually looks like when the agent runs autonomously — no confirmations, no hand-holding. Just code shipping itself.

AlexBuildsAi @AiandAlex · 2026-04-22

Stop

kd @Abdul1ahhh · 2026-04-22

Sahil Panhotra | Indie Builder | Dev @SahilPanhotra · 2026-04-22

i am loosing my mind

claude team gives updates daily

take some rest guys😅