展示HN:基于GitHub的原生预测市场(使用Issues,无需数据库)
我建立了一个微型预测市场,完全依赖于 GitHub Issues,没有外部数据库。每个市场都是关于仓库工作的具体问题,例如:“问题 #123 会在 2025 年 11 月 15 日之前关闭吗?”合作者可以购买 YES/NO 股份来表达他们的看法,价格(0 - 1)反映了团队的集体预测。这有助于规划、优先级排序以及使期望变得透明——而不涉及真实货币。
在技术实现方面:在问题中嵌入了签名的 JSON 快照(市场和账本),使用 LMSR 定价,基于序列的乐观并发控制并支持重试,以及仅限合作者的交易。
我还没有为该仓库部署公共实例——计划在整合初步反馈后进行。仓库链接: [https://github.com/philippnagel/gantt](https://github.com/philippnagel/gantt)
我希望能收到关于治理(每个用户的上限、利益冲突)、哪些问题最有用(截止日期、PR 合并、发布)以及基于评论的交易用户体验的反馈。
查看原文
I built a tiny prediction market that lives entirely in GitHub Issues—no external database. Each market is a concrete question about repo work, e.g., “Will Issue #123 be closed by 2025-11-15?” Collaborators buy YES/NO shares to express beliefs, and the price (0 - 1) reflects the team’s collective forecast. This helps with planning, prioritization, and making expectations transparent - without real money.<p>Under the hood: signed JSON snapshots (market and ledger) embedded in issues, LMSR pricing, seq-based optimistic concurrency with retries, and collaborator-only trading.<p>I haven’t deployed a public instance yet for the repo - planning to do so after incorporating initial feedback. Repo:
<a href="https://github.com/philippnagel/gantt" rel="nofollow">https://github.com/philippnagel/gantt</a><p>I’d love feedback on governance (caps per user, conflicts of interest), which questions are most useful (deadlines, PR merges, releases), and the UX of comment-driven trading.