What does this skill do, and when should you use it?
This skill provides change-size guidance for code review. It recommends keeping non-mechanical changes at or below 800 changed lines, with complex logic changes under 500 lines. When a change exceeds those limits, it evaluates whether the work can be split into reviewable stages. Its staging advice is based on the actual diff, dependencies, and affected call sites.
Counts or assesses the total changed lines in a proposed change and compares the result with the 800-line general guideline and 500-line complex-logic guideline. For oversized changes, it examines the actual diff, dependencies, and affected call sites, then explains whether staging is possible and identifies the smallest coherent first stage.
- A developer wants to assess a large non-mechanical change before opening a pull request.
- A reviewer needs a concrete size guideline for a change involving complex logic.
- A team is preparing to split a large submission into smaller review stages.
- A maintainer needs staging advice that accounts for dependencies and call-site impact.
How do you install this skill?
- This is a low-confidence static assessment based only on source files; the skill was not executed or independently tested.
- The 800-line and 500-line thresholds have no stated basis and should not replace judgment about risk, dependencies, call sites, or review capacity.
- No standardized output format is specified, so staging recommendations may be inconsistent.
- Local filesystem
The repository bundles 20 skills, and its README does not document separate installation for this skill. To install the Codex CLI collection, run curl -fsSL https://chatgpt.com/codex/install.sh | sh on macOS or Linux, or powershell -ExecutionPolicy ByPass -c "irm https://chatgpt.com/codex/install.ps1 | iex" on Windows. The README also lists npm install -g @openai/codex and brew install --cask codex.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/openai/codex.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/.codex/skills/code-review-change-size" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Invoke the skill during a code-review workflow with access to the relevant change. Ask Codex to assess the change size; if it exceeds the guidance, ask for staged delivery based on the diff, dependencies, and affected call sites. The supplied README and SKILL.md do not document a dedicated trigger command or example prompt.
What are this skill's strengths and limitations?
- Sets explicit 800-line and 500-line review-size guidelines.
- Requires staging advice to consider the actual diff, dependencies, and call sites.
- Has a narrow, understandable focus on reviewable change size.
- Does not define the exact method for counting changed lines.
- Provides no test suite, automated check, or integration command.
- Does not document standalone installation or invocation.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| Change Size Review Guide this page ✓ OpenAI · Official | 48 · Use with care | ★ 129k | 11d ago | Apache-2.0 |
| OpenChamber Change Discipline | 52 · Use with care | ★ 12k | 11d ago | MIT |
| Diff Risk Explainer | 14 · Not recommended | ★ 32k | 12d ago | Apache-2.0 |
| Codex Integration Skill (skill-codex) | 44 · Not recommended | ★ 1.5k | 1mo ago | MIT |
| Review Agent Code Review Skill ✓ OpenAI · Official | 52 · Use with care | ★ 129k | 11d ago | Apache-2.0 |
How did FollowSkills review this skill?
The skill only gives change-size guidance and declares no permissions, external data flows, dependencies, or destructive actions, so its risk surface is limited; however, it lacks user-confirmation, sensitive-data, rollback, and concrete attribution procedures, so 8 points are deducted.
The rules are concise and internally consistent: 800 lines for ordinary changes, 500 for complex logic, with staging guidance above the limit; however, there is no handling for empty diffs, unavailable dependency or call-site information, mechanical-change ambiguity, or abnormal-input feedback, so 12 points are deducted.
The intended use case—controlling change size during code review—is reasonably clear; however, metadata and instructions do not sufficiently define non-fit cases, input/output format, semantic trigger conditions, or Chinese-language usage, so 7 points are deducted.
The information architecture is clear, and repository-level Apache-2.0 licensing and official provenance are available; however, the skill lacks installation notes, examples, parameter definitions, limitation disclosure, versioning, changelog, maintenance ownership, and update path, so 9 points are deducted.
It can directly help reviewers decide when to split a large change and requires considering the actual diff, dependencies, and call sites; however, it provides no directly deliverable output template or verification procedure, and evidence for the thresholds is limited, so 9 points are deducted.
The skill file itself is auditable, and its rules and decision conditions are traceable; however, it contains no tests, example results, CI coverage, or third-party reproduction evidence, so only limited static-review credit is justified and 7 points are deducted.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →