--- title: "Reviewing AI-Generated Code" url: https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-ai-generated-code/ --- # Reviewing AI-Generated Code This document uses key words such as "MUST", "SHOULD", and "MAY" as defined in [RFC 2119](https://www.ietf.org/rfc/rfc2119.txt) to indicate requirement levels. Statuscandidate Version`1.0.0`[(changelog)](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-ai-generated-code.md#changelog) ## [Overview](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-ai-generated-code.md#overview) This playbook extends the standard code review process with AI-specific checks for common failure modes in AI-generated code. It covers hallucinated imports, meaningless tests, over-engineering, speculative changes, missing context, and subtle behavior changes. By following these steps, reviewers will catch issues that automated tools miss while maintaining the same quality standards as human-written code. Related resources: * [Reviewing a PR](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-a-pr.md) — base review process * [Code Quality Standards](https://develop.sentry.dev/sdk/getting-started/standards/code-quality.md) — test quality requirements * [Sentry Skills](https://github.com/getsentry/skills#available-skills) — find-bugs skill for systematic detection *** ## [Standard review first](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-ai-generated-code.md#standard-review-first) Apply the full review checklist from [Reviewing a PR](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-a-pr.md): #### [1. Check the PR description](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-ai-generated-code.md#1-check-the-pr-description) What, why, linked issue. #### [2. Check CI status](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-ai-generated-code.md#2-check-ci-status) You **MUST NOT** review failing code. #### [3. Review for common issues](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-ai-generated-code.md#3-review-for-common-issues) Runtime errors, performance, side effects, backwards compatibility, security, test coverage ([Test requirements by change type](https://develop.sentry.dev/sdk/getting-started/standards/code-quality.md#testing)), test quality ([Test quality](https://develop.sentry.dev/sdk/getting-started/standards/code-quality.md#testing)). #### [4. Check @sdk-seniors review triggers](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-ai-generated-code.md#4-check-sdk-seniors-review-triggers) Public API, dependencies, schema changes, security-sensitive code, frameworks. #### [5. Use LOGAF prefixes on feedback](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-ai-generated-code.md#5-use-logaf-prefixes-on-feedback) ([Review feedback conventions](https://develop.sentry.dev/sdk/getting-started/standards/review-ci.md#code-review)) #### [6. Approve when only `l:` items remain](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-ai-generated-code.md#6-approve-when-only-l-items-remain) *** ## [Additional AI-specific checks](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-ai-generated-code.md#additional-ai-specific-checks) AI-generated code has specific failure modes. You **MUST** check for these in addition to the standard review: #### [1. Hallucinated imports and APIs](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-ai-generated-code.md#1-hallucinated-imports-and-apis) Verify every import and function call actually exists. AI tools sometimes reference packages, modules, or functions that don't exist or have different signatures than expected. #### [2. Tests that test nothing](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-ai-generated-code.md#2-tests-that-test-nothing) You **MUST** check that test assertions would actually fail if the feature broke ([Test quality](https://develop.sentry.dev/sdk/getting-started/standards/code-quality.md#testing)). Watch for: hardcoded expected values that happen to match the output, `assert True` or equivalents, testing mock behavior instead of real behavior, asserting only that no exception was thrown. #### [3. Over-engineering](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-ai-generated-code.md#3-over-engineering) AI tools frequently add unnecessary abstractions, configuration options, and error handling for impossible cases. Ask: "does this need to be this complex?" If a simpler approach works, request it. #### [4. Speculative changes](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-ai-generated-code.md#4-speculative-changes) Code changes beyond what the issue or PR describes ([One logical change per PR](https://develop.sentry.dev/sdk/getting-started/standards/code-submission.md#one-logical-change-per-pr)). If the PR is "fix null check" but also reorganizes imports and adds docstrings, request a split. #### [5. Missing architecture context](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-ai-generated-code.md#5-missing-architecture-context) AI tools may not understand SDK-specific patterns and conventions. Check that the change fits the SDK's existing architecture, not just generic "good code" patterns. #### [6. Subtle behavior changes](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-ai-generated-code.md#6-subtle-behavior-changes) Pay extra attention to edge cases in any "cleanup" or "refactor" PR. AI refactors sometimes change semantics in ways that aren't obvious from a quick scan. You **SHOULD** use the [`sentry-skills:find-bugs`](https://github.com/getsentry/skills#available-skills) skill for systematic bug and vulnerability detection in the diff. ## [Referenced Standards](https://develop.sentry.dev/sdk/getting-started/playbooks/development/reviewing-ai-generated-code.md#referenced-standards) * [Review feedback conventions](https://develop.sentry.dev/sdk/getting-started/standards/review-ci.md#code-review) — LOGAF scale and blocking criteria * [Test requirements by change type](https://develop.sentry.dev/sdk/getting-started/standards/code-quality.md#testing) — test coverage expectations * [Test quality](https://develop.sentry.dev/sdk/getting-started/standards/code-quality.md#testing) — meaningful assertion requirements * [AI attribution](https://develop.sentry.dev/sdk/getting-started/standards/code-submission.md#ai-attribution) — Co-Authored-By footer requirement * [One logical change per PR](https://develop.sentry.dev/sdk/getting-started/standards/code-submission.md#one-logical-change-per-pr) — focused PR scope *** | Version | Date | Summary | | ------- | ---------- | ------------------------------------------------------------------------------------------------ | | `1.0.0` | 2026-02-21 | Initial playbook — specialized review techniques for AI-generated code with common failure modes |