Guide · Agent skill
How to use the Multi-Reviewer Patterns agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
Multi-Reviewer Patterns is an agent skill, a set of instructions that an AI coding agent reads when a task matches its description. It guides the agent through code reviews split across quality areas such as security, performance, architecture, testing and accessibility. This guide explains what the skill does, how to install it, how to ask for it, and where it fits best, so you can run a combined review and get one organized report.
Multi-Reviewer Patternswshobson/agents · MIT Download the skill Quick start
- 1Download the ZIP. Get the multi-reviewer-patterns ZIP from the skill page on this site.
- 2Unzip into a skills folder. Extract it into ~/.claude/skills/ for personal use, or into .claude/skills/ in a project to share it with that project.
- 3Or upload in the Claude apps. In the Claude apps, go to Settings → Capabilities → Skills and upload the ZIP.
- 4Describe a multi-reviewer task. Ask in plain language for a multi-dimension review, severity calibration or a consolidated report.
- 5Pick the review dimensions. Let the agent choose from Security, Performance, Architecture, Testing and Accessibility using the recommended combinations.
- 6Merge and rate findings. Ask the agent to apply the merge rules and severity criteria to the combined reviewer output.
- 7Check the consolidated report. Review the Critical-to-Low findings, the dimension summary and the recommendation before acting on them.
What the skill makes the agent do
Once installed, the agent follows a structured playbook for reviews that cover several quality dimensions. The skill includes a table of five dimensions (Security, Performance, Architecture, Testing and Accessibility), each with a focus and a guide to when it applies. For example, Security is marked as always relevant for code that handles user input or auth, and Accessibility is for UI and frontend changes.
The skill also covers three other jobs: choosing which dimensions to assign for a given change, merging duplicate findings from different reviewers, and rating severity with consistent criteria. The output follows a standard consolidated report template with findings grouped by severity and a summary table by dimension.
Install the skill
You can install the skill in two ways. For Claude Code, download the ZIP from the skill page and unzip it into ~/.claude/skills/ for personal use, or into .claude/skills/ inside a project to share it with that project. In the Claude apps, you can instead upload the ZIP under Settings → Capabilities → Skills.
The download contains SKILL.md and references/review-dimensions.md. Keep the folder structure intact so the agent can find the checklists.
Trigger the skill with natural requests
You do not need a special command. The skill's description says to use it when organizing multi-reviewer code reviews, calibrating finding severity, or consolidating review results, so ask for those things in plain language. Name the target and the kind of review you want, and the agent can pick up the skill from its description.
Good requests describe the change and the outcome, for example asking the agent to plan a multi-dimension review of a large pull request, to rate findings from several reviewer passes on one consistent scale, or to combine the results into one report.
- Organize a multi-dimension review of this pull request and tell me which areas each reviewer should cover.
- Merge these three review outputs, remove duplicate findings and rate severity consistently.
- Consolidate these review results into a single report with a summary table by dimension.
Use the merge and severity rules
When two reviewers flag the same file and line for the same issue, the skill says to merge them into one finding and credit all reviewers. If they describe different issues at the same spot, both stay as separate findings tagged as co-located. When severity conflicts, the higher rating is kept.
Severity uses four levels: Critical, High, Medium and Low, judged on impact and likelihood. The calibration rules state, for instance, that externally exploitable security vulnerabilities are always Critical or High, and that style issues with no functional impact are Low.
- Same file:line and same issue: merge and credit all reviewers.
- Same file:line but different issues: keep both as separate findings.
- Same issue in different locations: keep separate and cross-reference.
- Conflicting severity: use the higher rating.
- Conflicting recommendations: include both with reviewer attribution.
Read the consolidated report
The report template opens with the target, the reviewers used, the date and the file count. Findings are grouped under Critical, High, Medium and Low headings with IDs such as CR-001, and each finding lists its location, dimension, description, impact and fix. A summary table counts findings per dimension and severity, and a closing Recommendation section gives an overall assessment and prioritized actions.
## Code Review Report
**Target**: {files/PR/directory}
**Reviewers**: {dimension-1}, {dimension-2}, {dimension-3}Requirements and safety notes
The skill lists no requirements, runs no scripts and makes no network calls. It is only instructions and templates. It contains no analysis tools, so the findings come from the agent's own review work, not from automated scanning.
Because the severity guidance is generic, do not treat the output as a security audit or as sign-off on its own for security-critical code.
Tips for better results
Pick dimensions deliberately. The recommended combinations table gives a starting point, such as Security, Performance and Architecture for API endpoint changes. Adding every dimension to a small change adds overhead without much benefit.
Keep the reviewer outputs in one place before asking for consolidation, so the agent can compare the file and line references across them.
What people use it for
- Planning which specialist reviewers should cover a large pull request.
- Combining several AI review passes on the same code into one report.
- Making severity ratings consistent across reviewers working on the same release.
- Producing a uniform review report format for a team.
Frequently asked questions
How do I install the multi-reviewer-patterns skill?
Download the ZIP from the skill page and unzip it into ~/.claude/skills/ for personal use, or into .claude/skills/ in a project. In the Claude apps, you can upload the ZIP under Settings → Capabilities → Skills instead.
What does the multi-reviewer-patterns skill require?
The skill lists no requirements. It runs no scripts and makes no network calls, and it consists of SKILL.md and a references file with review checklists.
Is multi-reviewer-patterns free to use?
The skill is published under the MIT licence. The source does not mention pricing, so check the skill page for any terms beyond the licence.
Does the skill scan my code for bugs automatically?
No. It contains no analysis tools and relies on the agent's own review work. It provides a playbook, severity criteria and report templates rather than automated scanning.
When should I use this skill instead of a single review pass?
Use it for larger changes that span several areas, or when you need consistent severity ratings and a uniform report. For quick single-pass reviews, the overhead of multiple dimensions is usually unnecessary.
Can this skill replace a security audit?
No. Its severity guidance is generic, so it should not be the only sign-off for security-critical code. Use a qualified audit for that.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
This skill suits engineering leads running a review of a large change across several specialists, and teams that want consistent severity ratings and one report format. Skip it for quick single-pass reviews, for automated scanning, or as a standalone security sign-off, where a dedicated scanner or a qualified audit is the better choice.
Download the skill