Cmd. /code-review
Review a diff, PR, branch, or path (skill).
Written and reviewed by Eric Han · Co-founder & CTO, reopt
Key takeaways
- Review a diff, PR, branch, or path (skill).
- Check the input surface and availability before using the examples below.
Purpose
Review a diff, PR, branch, or path (skill).
/code-review [low|medium|high|xhigh|max|ultra] [--fix] [--comment] [--max-findings n|all|default] [pr#|branch|path]Availability
| Field | Value |
|---|---|
| Input surface | Running CLI prompt |
| Kind | Bundled skill |
| Status | Current reference |
| Aliases | /review |
| Documented minimum | Not specified |
| Reviewed against | 2026-10-23 · Claude Code 2.1.308 |
The installed command menu and account or organization settings determine availability. --max-findings requires v2.1.288+.
Examples
/code-review highEffects and persistence
Reviews the diff or selected target. --fix requests edits and --comment requests posted comments, expanding the operation's effects. A --max-findings value is reused by later reviews until you pass default.
Review execution
For models without tuned review settings, /code-review uses leaner inline prompts rather than many review subagents (v2.1.274). When no base branch or shared history exists, claude -p refuses the whole-repository fallback. A user running claude ultrareview directly consents to that fallback; an agent invoking it through a tool does not. Reviewing a first commit requires interactive confirmation (v2.1.277). From v2.1.290, medium effort also reports cleanup and CLAUDE.md-convention findings on models without tuned review settings, including Opus 5.5 and Sonnet 5.5.
Posting comments
/ultrareview --post posts a PR comment directly when findings arrive and prints its link, without a second cloud session. Use --post when you intend to publish findings to the PR.
official changelog · 2026-09-12 KST
Cautions
A plain review, --fix, and --comment have different effects. Check the selected target before requesting changes or publication. --max-findings <n>, all, or default (v2.1.288+) changes how many findings are reported, and later reviews reuse the value until you pass --max-findings default.