New to Claude Skills? Learn how to install them →

alibaba on GitHub

Open Code Review Delegate

Free

Enhance code reviews with agent-driven insights.

by alibaba20k stars on alibaba/open-code-review
1 views
Updated Aug 10, 2026
Get this skill

Free · Opens the source repo

What Open Code Review Delegate does

Open Code Review Delegate is a skill designed to streamline the code review process by leveraging the capabilities of the host agent to perform reviews. Instead of relying on an external LLM endpoint, this skill allows the host agent to take charge of the review process, utilizing the Open Code Review (OCR) framework for deterministic tasks such as file selection and rule resolution. This approach enables developers to harness the full potential of their AI tools while maintaining control over the review process.

The workflow begins with a preview step, where users can identify which files are eligible for review based on various criteria, including specific branches or commits. The skill outputs essential metadata, including the reviewable file list and any excluded files, allowing for a clear understanding of what will be reviewed. Following this, users can retrieve rules associated with the selected files, ensuring that the review adheres to established guidelines without redundancy.

Once the files and rules are established, users can proceed to examine the diffs of the code changes using standard Git commands. This flexibility allows for thorough reviews of each file, with the ability to categorize findings based on severity. The skill also allows for optional fixes to be applied directly for high or critical issues, enhancing the overall efficiency of the review process. The final step involves generating a comprehensive report that summarizes the review, ensuring accountability and clarity in the review outcomes.

This skill is particularly beneficial for teams looking to integrate AI into their code review workflows without losing the nuances of human oversight. By delegating the review tasks to the host agent, developers can focus on more complex aspects of their code while ensuring that standard practices are upheld throughout the review process.

When to use it

Use this skill when you want to leverage your host agent's capabilities to perform detailed code reviews with structured outputs.

When not to use it

This skill may not be suitable for environments where LLMs are essential for generating insights or where complex code analysis is required beyond basic rule checks.

What you can build with it

Reviewing a Feature Branch

Use the skill to compare a feature branch against the main branch, identifying changes and ensuring compliance with coding standards.

Conducting a Commit Review

Quickly review a specific commit by using the commit hash to focus on recent changes and their implications.

Evaluating Workspace Changes

Utilize the skill to assess uncommitted changes in your workspace, ensuring that all modifications are reviewed before integration.

How to install Open Code Review Delegate

View source

1. Install with the skills CLI

npx skills add alibaba/open-code-review/open-code-review-delegate --agent claude-code

2. Or install it manually

Download the skill folder and drop it into ~/.claude/skills/ for all projects, or .claude/skills/ to scope it to one repo. Restart Claude Code so it picks up the new skill.

Anthropic's agentic coding CLI, and the reference implementation of Agent Skills. Drop a skill folder into ~/.claude/skills and Claude Code loads it automatically whenever a task matches the skill's description. Claude Code docs

Inside SKILL.md

Written by alibaba

Open Code Review — Delegation Mode

This Codex plugin skill intentionally mirrors the canonical skill at skills/open-code-review-delegate/SKILL.md. Keep both files synchronized when updating OCR delegation instructions; a symlink is avoided because plugin installs may only materialize the plugin subtree.

A skill for performing AI code review where OCR provides deterministic engineering (file filtering, rule resolution) and the host agent performs the actual review using its own intelligence and tools.

Prerequisites

which ocr || echo "NOT INSTALLED"

If ocr is not installed:

npm install -g @alibaba-group/open-code-review

No LLM configuration is needed for delegation mode.

Workflow

Step 1: Preview — Determine What to Review

ocr delegate preview --format json [--from <ref> --to <ref>] [--commit <hash>] [--exclude <patterns>]

This outputs:

  • mode (workspace / range / commit)
  • from / to / commit / merge_base — ref metadata for constructing git commands
  • Reviewable file list — paths, status, insertions/deletions
  • Excluded files — with exclusion reason

Common invocations:

ScenarioCommand
Workspace changesocr delegate preview
Branch comparisonocr delegate preview --from main --to feature
Single commitocr delegate preview -c abc123

Step 2: Get Rules for Files

ocr delegate rule --format json <path1> <path2> ...

Pass the reviewable file paths from Step 1. Output is grouped by rule content — files sharing the same rule appear under one group, avoiding repetition.

Step 3: Get Diffs

Use git directly based on the mode/ref info from Step 1:

Range mode (merge_base provided in preview output):

git diff <merge_base>..<to> -- <path>

Commit mode:

git show <commit> -- <path>

Workspace mode:

# Tracked files
git diff HEAD -- <path>
# New untracked files — read directly (entire file is new code)
cat <path>

Step 4: Review Each File

Create a checklist containing every reviewable_files entry. For each reviewable file:

Use (path, status) as the checklist identity. Workspace mode can report the same path twice when a staged deletion is followed by an untracked recreation.

  1. Get its diff (Step 3)
  2. Consult its Rule Group (from Step 2) for the review checklist
  3. Conduct a thorough review, using appropriate context tools as needed
  4. Mark the file reviewed, or skipped with a concrete reason

For large changes, review in bounded batches grouped by shared rules and diff size. Do not stop after finding the first high-severity issue.

Step 5: Format Output

Each comment must follow this structure:

FieldTypeRequiredDescription
pathstringyesRelative file path
contentstringyesReview comment describing the issue
start_lineintegernoStart line in the new file
end_lineintegernoEnd line in the new file
categoryenumnobug, security, performance, maintainability, test, style, documentation, other
severityenumnocritical, high, medium, low

Step 6: Classify and Report

Before reporting, verify that every previewed file is accounted for. Include total_files, reviewed_files, skipped_files, and coverage_rate in the summary. A skipped file must include its reason.

Group findings by severity:

  • Critical/High: Bugs, security issues, data loss risks — always report
  • Medium: Performance concerns, error handling gaps, maintainability issues — report with context
  • Low: Style nits, minor suggestions — report only if clearly valuable

Discard likely false positives silently.

Step 7: Fix (Optional)

If the user requested "review and fix":

  • Apply High/Critical fixes directly
  • Describe Medium fixes that require manual intervention
  • Skip Low-priority items unless trivial

Sub-commands Reference

CommandPurpose
ocr delegate previewWhich files to review + mode/ref metadata
ocr delegate rule <path...>Review rules grouped by content

Shared Flags

FlagDescription
--from <ref>Source ref for range mode
--to <ref>Target ref for range mode
-c, --commit <hash>Single commit mode
--repo <path>Repository root (default: cwd)
--rule <path>Custom rule.json path
--exclude <patterns>Comma-separated exclude patterns
-b, --background <text>Business context
-B, --background-file <path>Business context from Markdown file
-f, --format <text|json>Output format; use json for agent integrations

Gotchas

  • No LLM needed on OCR side — delegation mode never calls an LLM. All intelligence comes from the host agent.
  • Rules are grouped — Files sharing the same rule are grouped together in the output. You can pass any number of paths per call; for large changes, fetch rules per-batch as you review.
  • Working directory mattersocr delegate operates on the Git repo at the current directory. Use --repo /path to override.
  • Untracked files in workspace modepreview includes untracked files. For these, read the file directly instead of using git diff.
  • Background context — pass --background to preview when you have requirement context; it appears in the output for your reference during review.
  • Coverage is mandatory — every reviewable_files entry must end as reviewed or explicitly skipped; do not silently omit files.

Frequently asked questions about Open Code Review Delegate

Similar skills