Best of Code Review — March 2026

  1. 1
    Article
    Avatar of addyAddy Osmani·9w

    Comprehension Debt — the hidden cost of AI generated code.

    Comprehension debt describes the growing gap between how much code exists in a system and how much any human genuinely understands. As AI coding tools generate code faster than engineers can meaningfully review it, teams accumulate invisible risk: tests pass, velocity metrics look healthy, but no one can explain why design decisions were made or how parts of the system interact. An Anthropic study found engineers using AI assistance scored 17% lower on comprehension tests than those who didn't. Tests and detailed specs help but don't fully solve the problem—tests can't cover behavior no one thought to specify, and specs can't capture all implicit implementation decisions. The real scarce resource becomes engineers who deeply understand the system. Comprehension debt is more insidious than technical debt because it accumulates invisibly, nothing in standard measurement systems captures it, and the reckoning arrives at the worst possible moment. The solution is treating genuine understanding—not just passing tests—as a non-negotiable part of shipping software.

  2. 2
    Article
    Avatar of francofernandoThe Polymathic Engineer·7w

    What Really Makes a Succesful Software Engineer

    Technical skills get you hired, but five non-technical traits determine long-term success as a software engineer: patience (debugging is inherently slow and iterative), determination (roadblocks are inevitable and quitting is always tempting), a student mindset (the industry never stops changing so neither should you), accepting criticism constructively (code reviews and feedback are growth tools, not attacks), and communication (explaining ideas clearly to humans is as important as writing logic for machines). These meta-skills underpin all learning and career progression, yet no course teaches them.

  3. 3
    Article
    Avatar of freecodecampfreeCodeCamp·11w

    OSS Pull Request Therapy: Learning to Enjoy Code Reviews with npmx

    A developer shares her personal journey from OSS skeptic to contributor on the npmx project, detailing her struggles with code review anxiety and perfectionism. Through her first real PR experience on npmx, she learned to embrace collaboration over perfection and developed a more nuanced view of open source. The post includes practical tips for evaluating OSS projects before contributing and advice for both PR authors and reviewers on building a healthier review culture.

  4. 4
    Article
    Avatar of agents_digestAgentic Digest·9w

    Claude Code gets 1M context for free, GitHub pulls premium models from student Copilot

    Anthropic silently expanded Claude Opus 4.6 and Sonnet 4.6 to support 1M token context by default at no extra API cost, removing a key constraint for Claude Code users working with large codebases. GitHub moved in the opposite direction, stripping premium models (GPT-5.4, Claude Opus 4.6, Sonnet 4.6) from its free Copilot Student plan citing sustainability, drawing nearly 2,900 downvotes. A live benchmark of 22 code review tools ranked Claude first on quality but last on cost at $23.60 per review — roughly 1,100x more expensive than the most efficient tool. NanoClaw, an open-source agent framework endorsed by Andrej Karpathy with 22K GitHub stars, formalized a Docker partnership to run agents in isolated MicroVM sandboxes. Other notable items include Chrome v146 shipping native MCP support, shadcn/cli v4 with coding agent context features, AWS SAM integration for the Kiro IDE, and a documented case of an AI agent autonomously publishing a blog post attacking a maintainer who rejected its PR.

  5. 5
    Article
    Avatar of jetbrainsJetBrains·9w

    How We Taught AI Agents to See the Bigger Picture

    AI agents writing code for large legacy codebases tend to repeat outdated patterns because they mistake frequency for correctness. The JetBrains TeamCity team tackled this by building CommitAtlas, an internal tool that mines Git history to extract accepted patterns, naming conventions, and migration examples. Before writing code, agents query CommitAtlas for task-specific guidance derived from real, reviewed commits rather than raw codebase frequency. This approach reduced pull request rejections and helped agents produce code that fits the project's evolving standards rather than its historical defaults. The core insight: repository history is implicit documentation, and giving agents access to it bridges the gap between technically correct and genuinely acceptable code.

  6. 6
    Article
    Avatar of nolanlawsonRead the Tea Leaves·8w

    The diminished art of coding

    A reflection on how AI coding agents are transforming programming from a craft into an assembly-line process. The author argues that LLMs have resolved the tension between code-as-art and code-as-function firmly in favor of function, shifting developer focus from low-level elegance to high-level architecture. The post encourages developers to seek artistic fulfillment outside of coding — through painting, music, dance, or fiction — as the human touch diminishes in software creation. It closes with the observation that we're in a 'fast-fashion era' of coding: software is vibe-coded, used, and discarded.

  7. 7
    Article
    Avatar of github_updatesGitHub Changelog·10w

    Request Copilot code review from GitHub CLI

    GitHub CLI v2.88.0 now lets you request a GitHub Copilot code review directly from the terminal, without switching to the browser. You can add Copilot as a reviewer non-interactively with `gh pr edit --add-reviewer @copilot` or through interactive prompts when using `gh pr create` or `gh pr edit`. The release also improves reviewer and assignee selection with a search-as-you-type experience, replacing the previous approach of loading all collaborators upfront — boosting performance for large organizations and fixing accessibility issues with screen readers.

See all Code Review archives