Stop being the code review bottleneck - by Jina Yoon

Product for Engineers

SubscribeSign in

Stop being the code review bottleneck<br>4 ways to make AI code review suck less (with prompts)

Jina Yoon<br>Jul 09, 2026

32

Share

Agents are writing code faster than any human can review.<br>The naive solution would be for developers to review code faster. The 500 IQ take is for developers to review as little code as possible.<br>If you need to be involved in every code review, you will always be the bottleneck. Instead, put yourself outside of the code review loop by building a pipeline that delegates tasks to agents.<br>We asked engineers at PostHog how they’ve been reviewing AI-generated code to keep shipping fast without losing quality.<br>Here are four workflow changes you can steal (with prompts included) to make your life easier.

1. Make agents review code for you

The number one thing to add, if you haven’t yet, is a way for agents to review code for you.

The goal is to offload the simpler reviews to agents, and flag if something genuinely needs a human.<br>The key is that the agent that wrote the code can’t be the one that reviews it. Agents are bad at checking their own work since they’re often unaware of their own blind spots.1<br>For the same reason, it’s better to have multiple agents with different instructions and goals to cover more gaps,2 as well as different models and providers for different reviewers.3<br>Here’s how one of our engineers, Paul D’Ambra, makes his own custom agent review system work:

First, qa-swarm spawns four reviewer agents, each with their own special instructions:<br>qa-team – spawns technical subagents that hunt for security, database, performance, etc.

security-audit – probes for vulnerabilities like SQL or prompt injections

paul-reviewer – uses Paul’s voice and focuses on observability, rollouts, naming

xp-reviewer – applies an Extreme Programming lens to review

Then, review-triage sorts those reviews to classify threads into three categories:<br>actionable → gets fixed and pushed

nits → get resolved and replied to with a comment

ambiguous → escalated and sorted for Paul to work through with the agent later

An outer loop iterates up to three times or until no new actionable threads appear.

From there, you can connect this to another loop that shepherds the PR until it’s ready to merge – more on that in the next section.<br>The takeaway: Save time reviewing code by making agents review each other. This knocks out easier reviews so that only the PRs that really need human attention get flagged.

Steal this

You can check out and copy Paul’s qa-swarm and review-triage skills, or use this prompt to design your own review loop based on his:<br>Read Paul D'Ambra's qa-swarm skill, plus its sibling review-triage in the same folder, then help me design my own version: https://github.com/pauldambra/dotfiles/blob/main/ai/skills/qa-swarm/SKILL.md

It should take in a single PR, spawn a reviewer panel, triage every finding and existing PR thread into actionable / nit / ambiguous, and keep going until nothing's left but the ambiguous ones flagged for me.

Interview me about my stack, tooling, available models, and how autonomous it should be — what gets auto-fixed vs. only reported, and what it may post to GitHub — before writing the final SKILL.md, then install it.

That said, these systems can get token expensive:<br>“Something like 60% of my token spend is burned automating the toil of handling CI and review and I don’t regret a single dollar.” – Paul

So if running multiple agents or loops isn’t an option for your team, look for single agent designs like this one by Kun Chen.

2. Delegate PR babysitting to loops

The context switching that comes with agentic coding is exhausting. One easy way to reduce that fatigue is by automating code review-adjacent tasks that don’t need your attention.<br>For example, babysitting a single PR can involve tedious tasks like monitoring CI, re-running flaky tests, checking notifications for comments, and keeping the branch up to date.<br>Why waste your most precious resource – your energy – when you can just delegate all of it to a loop?<br>The takeaway: Reduce context switching and fatigue by delegating simple tasks like PR babysitting to a loop.

Steal this

You can implement your own PR babysitter skill, based on babysit-prs by Phil Haack with the prompt below. (It works best if you run it after creating a review loop skill from the previous section.)<br>Read https://github.com/haacked/dotfiles/blob/main/ai/skills/babysit-prs/SKILL.md and adapt it for me: same sweep/state design, but it dispatches my own single-PR review skill via a spawned agent per unreviewed PR.

Before writing SKILL.md, interview me on: which skill it dispatches and where my skills live, my stack/tooling/models, and which extra tasks to include — CI monitoring, branch freshness, flaky-test reruns, lint/format autofix, regenerating drifted artifacts, description sync.

Ground the interview in facts you can...