# How to Kill the Code Review — Ankit Jain, Aviator

Source: [AI Engineer](https://www.youtube.com/watch?v=YgEv7IQzGdM)  
Feed7 permalink: https://feed7.dev/p/how-to-kill-the-code-review-ankit-jain-aviator-0rku6kj  
Published: 2026-08-17T20:00:08.000Z  
Trust: Source Linked (source_linked)

## Why Included

As agent-written code overwhelms line-by-line review, preserve session decisions, turn recurring review comments into guardrails, and review verification evidence instead of diffs alone.

## Source Summary

More than **30% of changes** now merge without review, while reviewed changes reportedly wait **4× longer**. The proposed alternative separates alignment from semantic checking: retain agent-session decisions, derive acceptance criteria, combine them with recurring invariants, and verify a generated test plan.

## Practical Implication

Keep prompts and human corrections as review artifacts. Mine repeated comments into an **AI slop registry**, then have reviewers inspect intent, architecture, test plans, previews, and evidence rather than trusting an AI-reviewed diff.

## Agent-Ready Context

More than **30% of changes** now merge without review, while reviewed changes reportedly wait **4× longer**. The proposed alternative separates alignment from semantic checking: retain agent-session decisions, derive acceptance criteria, combine them with recurring invariants, and verify a generated test plan.

Keep prompts and human corrections as review artifacts. Mine repeated comments into an **AI slop registry**, then have reviewers inspect intent, architecture, test plans, previews, and evidence rather than trusting an AI-reviewed diff.

This is aimed at collaborating teams, not solo projects. Building the registry has an upfront cost, generated tests can still inherit blind spots, and some checks require an LLM fallback rather than deterministic verification.

## Connected Context

Feed7 judgment across 479 accumulated Signals:

This reframes code review as verification of retained intent, acceptance criteria, test plans, and evidence rather than line-by-line diff inspection. Against prior calls for bounded human review, it narrows what humans should review and adds a team memory mechanism for recurring failure patterns, while preserving human judgment where generated checks remain fallible.

- [Loop Engineering from First Principles — Kyle Mistele, HumanLayer](https://feed7.dev/p/loop-engineering-from-first-principles-kyle-mistele-humanlayer-1nuq7gf) — Loop Engineering pauses for human review of bounded PRs; this Signal changes the review object from the diff toward intent, test plans, previews, and evidence.
- [MineValiCoder: Reliable Code Generation with Test Case Quality Mining and Bipartite Graph-Based Mutual Validation](https://feed7.dev/p/2607-22471v1-0kivszm) — MineValiCoder directly addresses the stated risk that generated tests inherit blind spots by validating generated code and tests together.
- [Harness Engineering is not Enough: Why Software Factories Fail — Dex Horthy, HumanLayer](https://feed7.dev/p/harness-engineering-is-not-enough-why-software-factories-fail-dex-horthy-1tj1amt) — Both move alignment earlier, but this Signal proposes replacing much diff review with artifact verification, while the candidate warns against automating review away and retains human code ownership.
- [addyosmani/agent-skills](https://feed7.dev/p/agent-skills-0d3lx1j) — The skill pack’s evidence gates provide an implementation form for the proposed acceptance-criteria and test-plan review workflow, while the slop registry adds organization-specific learning from repeated comments.

## Context Map

- Layer: agent
- Domains: coding
- Topics: coding-agents, harness-engineering, agent-reliability

## Uncertainty

- This is aimed at collaborating teams, not solo projects. Building the registry has an upfront cost, generated tests can still inherit blind spots, and some checks require an LLM fallback rather than deterministic verification.

## Agent Instruction

Use this item as source-backed context. Do not invent claims beyond the linked source. If this item conflicts with another source, call out the conflict.