跳至正文
Hacker News · AI·· 44 分鐘前AI 評分58

Bugsworth releases AdversaryPR, a Claude skill that drafts pull request reviews with fixes

Show HN: AdversaryPR – A Claude skill to keep up with AI-written PR's

此內容暫無English版本,目前顯示的是English版本。

AI 導讀

Bugsworth, LLC released AdversaryPR 2.19.0, a free MIT-licensed Claude skill that reviews Azure DevOps and GitHub pull requests and drafts inline comments, each naming a failure scenario and showing a suggested fix as code. It reads the full diff plus callers and repo instruction files, runs one background review agent per PR in Claude Code, and posts nothing until the user approves selected comments; on GitHub an exact-line fix can post as a suggested change with a one-click commit. The skill never approves, merges or blocks a PR, does not run code or tests, and the author states there is no measured improvement data yet.

來源:Hacker News · AI · github.com