How does it handle different review environments?

It automatically detects if it is running in GitHub Actions to update a sticky comment or in a local environment to output a 'review-summary.md' file.

In what order are code findings presented?

Findings are grouped by severity in a specific order: Critical, Important, Debt, Suggested, and finally Questions regarding metadata.

When should I use the Posting Review Summary skill?

This skill should be invoked exactly once as the final step of a code review, after all specific inline comments have been posted to the files.

What is considered a deficient PR title or description?

It flags titles that are too generic (like 'fix bug') or descriptions that are empty or lack required screenshots for UI changes.

Does it create a new comment for every review update?

No, in GitHub environments it is designed to update a single sticky comment to prevent cluttering the PR conversation timeline.

Automated Code Review Summary

Name: Automated Code Review Summary
Author: bitwarden

bybitwarden

•

コラボレーションツール

Generates structured, severity-based code review summaries for GitHub and local environments to streamline pull request feedback.

The Posting Review Summary skill automates the final phase of the code review process by synthesizing all findings into a concise, scannable summary. It intelligently detects its environment to either update a sticky comment in GitHub Actions or generate a local markdown file, ensuring a consistent feedback loop. By enforcing a strict hierarchy of findings—ranging from critical security risks to minor suggestions—and assessing the quality of PR metadata like titles and test plans, it helps engineering teams maintain high code standards while reducing the manual overhead of formatting review conclusions.

主な機能

01Pull Request metadata assessment for titles and test plans

02Severity-based finding categorization (Critical to Suggested)

03Automated routing to avoid redundant PR comment notification noise

04Standardized markdown formatting with collapsible details

051 GitHub stars

06Automatic context detection for GitHub sticky comments or local files

ユースケース

01Creating local review artifacts during offline code audits and development.

02Enforcing PR documentation standards by flagging missing test plans or vague titles.

03Generating a final executive summary after completing a multi-file code review.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add bitwarden/ai-plugins posting-review-summary

For use in Claude.ai and ChatGPT

Download Skill