zaxiu88888/issue-rag

RAG over GitHub issues, with a hand-reviewed eval set and an audit of the judge itself. Reports the interval, not just the best number.

Python

335

1 contributors

MIT License

Versioned graph database on object storage. Git-style workflows and multimodal retrieval

Control what AI can do before it acts

Can AI agents reverse-engineer real websites?

Fastest Enterprise AI Gateway for Routing, Governing, and Securing all your AI Traffic

Momentum

How fast zaxiu88888/issue-rag is gaining stars, and where it lands on Trendshift and GitHub Trending

Stars, forks and merged pull requests each day or month, lined up with the repository's rank on Trendshift Trending

Nov 8 – Nov 12, 2026

Stars +87 in 5 days

Forks +3 in 5 days

Merged PRs 0 in 5 days

Trendshift Trendingwhere the repository ranked each day it made the top 25

Not in the Trendshift top 25 in this period

Nov 8Nov 9Nov 10Nov 11Nov 12

Stars, forks, merged pull requests and Trendshift Trending rank, by day
Day	Stars gained	Forks gained
Sun, Nov 8, 2026	+0	+1
Mon, Nov 9, 2026	+0	+1
Tue, Nov 10, 2026	+0	+1
Wed, Nov 11, 2026	+41	+0
Thu, Nov 12, 2026	+46	+0
Total	+87	+3

Recent discussions about this repository across the web

Nov 111 post
- 做 RAG 的人评测时，大多只报一个最好看的准确率。有人拿 vite 的 GitHub issue 做了个问答助手，连打分的模型也一起审了一遍。结果挺反直觉：加上检索以后，回答被判为编造的比例从 1.1% 涨到了 20%。作者的解释是：不检索时，模型会老实说不知道；加了检索，就变成很自信地答错。测试用了 95 条人工复核过的问题，生成和打分都用 deepseek-chat，检索语料 5473…
  Read full on X, posted by @xqliu
  @xqliu · x.com