PluginBench
Skill
Review
Audit score 70

git-summarize-weekly

zc277584121/mygitplugin

Summarize your weekly GitHub contributions for writing weekly reports

What is git-summarize-weekly?

Automatically generates English summaries of your GitHub contributions from the past week, organized by repository. Use this when preparing weekly reports or reviewing your work activity.

  • Calculates the date range for the current week (Monday to today)
  • Fetches recent GitHub events via the GitHub CLI API
  • Filters events to the current week and excludes your own repositories by default
  • Groups contributions by repository and event type (pushes, PRs, reviews, issues, comments)
  • Outputs a formatted English summary ready to paste into weekly reports

How to install git-summarize-weekly

npx skills add https://github.com/zc277584121/mygitplugin --skill git-summarize-weekly
Prerequisites
  • GitHub CLI (gh) installed and authenticated
  • Bash shell access
Claude Code
Cursor
Windsurf
Cline

How to use git-summarize-weekly

  1. 1.Request the skill to summarize your weekly GitHub contributions
  2. 2.The skill calculates the date range from Monday to today
  3. 3.It fetches your recent GitHub events using the GitHub API
  4. 4.Events are filtered to the current week and grouped by repository
  5. 5.Review the generated English summary and copy it into your weekly report

Use cases

Good for
  • Generate a weekly report summary for team standups or retrospectives
  • Track your open-source contributions across multiple repositories
  • Document your work activity for performance reviews or progress tracking
  • Quickly review which projects you contributed to during the week
Who it's for
  • Developers contributing to open-source projects
  • Team members writing weekly status reports
  • Engineers tracking their contribution history

git-summarize-weekly FAQ

Does it include contributions to my own repositories?

By default, it excludes your own repositories and only shows contributions to others' projects. You can explicitly request to include your own repositories if needed.

How far back does it look?

The GitHub Events API provides data for the past 90 days with up to 300 events. The skill filters this to only the current week.

What types of contributions are included?

It tracks pushes, pull requests, PR reviews, issues, issue comments, and branch/tag creation events.

What if I had very few contributions this week?

The skill will accurately report minimal activity if that's the case—it outputs what actually happened.

Can I customize the output format?

The skill outputs a standard format grouped by repository with brief descriptions, designed to be directly usable in weekly reports.

Full instructions (SKILL.md)

Source of truth, from zc277584121/mygitplugin.


name: git-summarize-weekly description: Summarize weekly GitHub contributions for writing weekly reports allowed-tools:

  • Bash

Git Summarize Weekly

总结本周的 GitHub 贡献,输出适合写入周报的英文摘要。

触发条件

当用户要求总结本周 GitHub 贡献、写周报或查看本周工作时使用此 skill。

执行步骤

  1. 计算日期范围:计算本周一的日期到今天。使用 date 命令计算。

  2. 拉取 GitHub events:

    gh api users/zc277584121/events --paginate --jq '.[]'
    

    获取用户近期的 GitHub 活动事件。

  3. 过滤和分析:

    • 过滤出本周时间范围内的事件。
    • 默认排除 zc277584121 自己账号下的仓库,只保留对他人仓库的贡献。
    • 如果用户明确要求包含自己的仓库,则不排除。
    • 关注以下事件类型:
      • PushEvent:代码推送
      • PullRequestEvent:PR 创建/合并
      • PullRequestReviewEvent:PR Review
      • IssuesEvent:Issue 创建/关闭
      • IssueCommentEvent:Issue 评论
      • CreateEvent:分支/标签创建
  4. 输出格式:

    • 必须用英文输出,适合直接抄写到周报中。
    • 大的 feature 或重要 PR 单独一条列出,简要说明做了什么。
    • 小的修复、review、评论等可以合并为一条。
    • 按仓库分组,格式简洁清晰。
    • 示例格式:
      ## Weekly Contributions (2025-01-20 ~ 2025-01-24)
      
      ### milvus-io/pymilvus
      - Implemented hybrid search API with support for multiple vector fields (#xxx)
      - Fixed connection pool timeout issue (#xxx)
      - Reviewed 3 PRs related to batch insert optimization
      
      ### milvus-io/milvus
      - Minor: commented on 2 issues about memory management
      

注意事项

  • GitHub Events API 最多返回最近 90 天、300 个事件的数据。
  • 如果某一周活动特别少,如实报告即可。
  • 输出要简洁实用,直接可以粘贴到周报中。