SetuHQ transit search

Name: SetuHQ transit search
Author: ashfordeOU

作者：ashfordeOU版本 0.1.0i★ –i在你的电脑上运行

注册名称io.github.ashfordeOU/setuhq

安装

概览

作者提供的说明（英文原文）

Search, vet and explain TESS transit candidates with provenance. Candidates, never confirmations.

PopNx MCP 的三大支柱

这三项永远不会合并成一个分数：意见、测量和人气是不同的东西。

用户评价i暂无评价评价功能将在后续阶段与账户功能一起开放
我们的基础检查i1 项中通过 1 项依据公开信息检查；尚未扫描代码
使用情况i尚未统计★ –

它能帮你做什么

我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。

连接方式: 在你的电脑上运行
传输方式: stdio
分类: 搜索与 AI 记忆

安装

选择你使用的 AI 应用，然后复制设置。

安装之前

基础检查：1 项中通过 1 项
在你的电脑上运行，拥有你的用户账户的权限（文件、网络、应用）。
需要你自己填写： SETUHQ_ROOT

你需要

PyPI已安装 Python 的 uv（用于 uvx 命令）

你用的是哪个应用？

ClaudeClaude CodeCursorVS CodeChatGPT

在上方选择你的应用。我们会记住它，下次按钮会显示“安装到 …”。

打开 Claude Desktop → Settings → Developer → Edit Config，把下面的内容粘贴到 claude_desktop_config.json

{
  "mcpServers": {
    "setuhq": {
      "command": "uvx",
      "args": [
        "setuhq",
        "mcp"
      ],
      "env": {
        "SETUHQ_ROOT": "your-value"
      }
    }
  }
}

在终端中运行这条命令

claude mcp add setuhq --env SETUHQ_ROOT=your-value -- uvx setuhq mcp

运行前请先填写空着的值。

在 Cursor 中打开

或者自己复制设置

粘贴到项目中的 .cursor/mcp.json，或 ~/.cursor/mcp.json

{
  "mcpServers": {
    "setuhq": {
      "command": "uvx",
      "args": [
        "setuhq",
        "mcp"
      ],
      "env": {
        "SETUHQ_ROOT": "your-value"
      }
    }
  }
}

浏览器会先询问是否打开应用，然后应用会显示设置并请你确认。密钥的值会留空，由你自己填写。

在 VS Code 中打开

或者自己复制设置

粘贴到项目中的 .vscode/mcp.json

{
  "servers": {
    "setuhq": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "setuhq",
        "mcp"
      ],
      "env": {
        "SETUHQ_ROOT": "your-value"
      }
    }
  }
}

浏览器会先询问是否打开应用，然后应用会显示设置并请你确认。密钥的值会留空，由你自己填写。

这个服务器在你的电脑上运行，而 ChatGPT 只能连接互联网上的服务器。请选择其他应用。

你的选择通过 Cookie 保存在这台设备上，无需账户。

需要你填写的值

SETUHQ_ROOT env
Data root the tools read and write (a SetuHQ checkout holds the stored runs, the catalogue and the benchmarks). No tool reaches outside it. Default: the checkout the package runs from, else the worki…

这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同，请以 README 为准。

它能做什么

它在你的电脑上运行；安装后，工具列表会显示在你的 AI 应用中。

基础检查i

我们检查的是任何人都能看到的信息；我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月7日

✓ 通过
作者与仓库一致
注册名称已通过 GitHub 账户 ashfordeOU 验证，该账户拥有此仓库
… 尚未检查
仍在维护
我们还没有从 GitHub 读取这个仓库
… 尚未检查
有许可证
我们还没有从 GitHub 读取这个仓库
– 不适用
端点有响应
它在你的电脑上运行，没有互联网端点
– 不适用
使用 OAuth 登录
在你的电脑上运行，改用本地设置
… 尚未检查
工具描述中没有隐藏指令
它在你的电脑上运行；我们只能读取互联网端点的工具列表

与“搜索与 AI 记忆”中的其他服务器比较i

星标数多于此分类中其他 7284 个服务器的 0%
发布到注册中心的时间比此分类中其他 7284 个服务器的 71% 更新
基础检查结果等于或优于此分类中其他 7284 个服务器的 100%

与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时，星标数才计入。

常见问题

SetuHQ transit search 安全吗？

它在 1 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前，请先看看这个服务器要求哪些权限，并使用只具备所需权限的密钥。

它需要哪些密钥或代码？

注册中心没有列出必需的密钥，不过有些服务器会在首次使用时要求你登录。

它在我的电脑上运行还是在云端？

它在你的电脑上运行。你的 AI 应用会在需要时启动它，数据不会经过作者的服务器（除非它自己调用在线服务）。

它还在维护吗？

GitHub 仓库最后更新于–，最新版本（0.1.0）于16天前进入注册中心。

哪些应用可以使用它？

支持 MCP 的应用，例如 Claude Desktop、Claude Code、Cursor 和 VS Code。你的电脑上还需要有 Python (uv)。

有什么可以替代它？

同一分类中有：Context7, Codebase Memory, Exa

替代选择

同一分类中其他作者的服务器，按星标数排序。

如何看卡片：★ GitHub 星标i三大支柱：评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i

Context7

作者：upstash

Up-to-date docs for AI agents

电脑或云端★ 64,224MIT昨天更新

安装

Codebase Memory

作者：DeusData

Codebase knowledge graph for AI agents — 162 languages, sub-ms queries, 99% fewer tokens.

在你的电脑上运行★ 48,148MIT21天前更新

安装

Exa

作者：exa-labs

Connect AI agents to Exa for web search, content fetching, and multi-step research.

在云端运行★ 5,213MIT上个月更新

安装

作者的 README

打开 README

按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。

SetuHQ · सेतु · the bridge

The open exoplanet observatory: from public sky data to reviewed, explained candidates.

setuhq.com · The app · How it works · Evidence · Quick start · Editions · Cite · Licence

Every dip in the sky, reviewed and explained

TESS has watched about 98% of the sky. Almost all of that light is public, and almost none of it has been looked at dip by dip. SetuHQ searches any TESS star: give it a TIC id and a sector or two, and it does the looking: a reviewer panel of AI Sentinels and deterministic Probes marks every dip in blind-coded light curves, the way Planet Hunters volunteers do. Each verdict keeps its marks, its weights and, for the Sentinels, their reasons in words. Then the pipeline does the physics: limb-darkened transit fits, bounded false-alarm probabilities, odd/even and secondary checks, and an honest label on everything.

*Sentinels run on Claude (Anthropic) by default; other providers are supported and labelled. Probes are deterministic matched-filter detectors.*

What SetuHQ is built to be:

A search you can audit. Every candidate carries its provenance: data hash, config hash, code fingerprint, every reviewer's marks.
A benchmark that keeps everyone honest, including SetuHQ. A frozen, blind test on 662 real TESS candidates, with controls, a public leaderboard and its negative results printed first.
An observatory you can explore. Every known exoplanet in 3D, a full catalogue, and fifteen animated chapters on how planets are found, all in one app.

Worked examples

Five stored runs on public TESS light curves from MAST, chosen to show different kinds of star and signal. Each keeps its data hash, configuration hash and code fingerprint; every dip is counted by its vetting outcome (a candidate stays unconfirmed; a rejected dip is not a candidate). Known planets are named with their reference: finding one again checks the pipeline, it is not a discovery. Every run but pi Men is Probes only; Sentinels not run on this example.

Star	Kind	TESS sector	Dips	Against the literature
WASP-126	Hot Jupiter	1	8 dips (8 candidates)	WASP-126 b: 8 of 8 transits (8 candidates).
TOI-270	Multi-planet system	5	7 dips (5 candidates, 2 flagged)	TOI-270 b: none of 6 transits; TOI-270 c: 5 of 5 transits (5 candidates); TOI-270 d: 2 of 2 transits (2 flagged; no period).
HD 15337	Small planets near the limit	4	3 dips (2 candidates, 1 flagged)	HD 15337 b: 1 of 5 transits (1 candidate; BLS period not the planet's); HD 15337 c: 2 of 2 transits (1 candidate, 1 flagged; BLS period not the planet's).
TOI-236	Known false positive	2	6 dips (6 rejected)	Caught: all 6 dips on the TOI ephemeris were rejected by vetting (secondary_any_phase, secondary_eclipse), so none is a candidate.
pi Men	Reference check	1	7 dips (5 candidates, 1 flagged, 1 rejected)	Rp/R★ and T14 within 5% of Huang et al. 2018 (docs/results/pi_men_validation.md).

Why each was chosen, its full comparison with the literature and the one command that reproduces it: docs/SHOWCASE.md. On the site, each has its dips, case file, plates and run report.

Their known planets to one scale, beside Earth, Neptune and Jupiter: artist's impressions chosen from each planet's measured radius and computed irradiation, drawn by the site's own renderer; every radius is exoplanet.eu's.

What's new in 0.1.0

The first public release (CHANGELOG). Nothing on this page is a discovery claim: SetuHQ finds and vets *candidates*.

The name. The project developed as *exoplan* is now SetuHQ, *setu* being Sanskrit for bridge: from public sky data to reviewed candidates. The package and the command are setuhq, the MCP server setuhq-mcp; the old exoplan commands still run, with a deprecation notice. The logo's Q is a star with a planet crossing its face (brand kit).
Open licence. AGPL-3.0-only, or a commercial licence from Ashforde OÜ; data, benchmark, leaderboard data and paper under CC BY 4.0 (Licence).
The feed's pre-registered rule. The search feed now needs a signal in both reviewed sectors and a FAP below 0.001. On held-out data, measured on the current code, it cuts noise about 4.5-fold but still admits about 6 false signals per 100 targets, so its admissions are leads for review (see Evidence).
In this release: pixel-level and Data-Validation-style vetting,

只显示了开头部分；其余内容请在仓库中阅读