Research Radar

English · 简体中文

A customizable static website for following research papers, publishing detailed reading notes, and keeping a searchable archive. Adapt the research scope to your discipline; add a dataset or database directory only if it helps your work.

Built with Astro, Markdown, JSON, and build-time KaTeX. No backend database, paid API, or AI account is required to run the website.

Live example

Genetic Daily Papers — papers.lucajiang.com is the original research website from which this template was extracted. Browse it for an example of detailed paper notes, dated archives, topic navigation, search, and a research resource directory in everyday use.

The example site is customized for statistical genetics and related methods. Its research scope, paper collection, and QTL-specific resource fields are personal to that site. Your copy starts with neutral examples, configurable topics, and the resource directory disabled; adapt these to your own field.

Start here

1. Create your own repository. Click Use this template → Create a new repository if the template button is available; otherwise Fork this repository. Use a template for a fresh history; fork if you want the upstream relationship. Do not clone this project's URL and then try to push your personal content here.
2. Choose a deployment path. Cloudflare Pages is the recommended starting point for Git-connected hosting at a root URL. GitHub Pages keeps hosting with GitHub and includes a ready-to-use workflow. Vercel and Netlify are alternatives.
3. Customize your site. Edit site.config.json, src/data/topics.json, and src/data/research-profile.json. The two published demonstration files are clearly fictional: replace or remove them before your first real issue.
4. Verify and publish. Run the commands below, then follow the chosen deployment guide.

# Replace YOUR-USERNAME and YOUR-REPOSITORY with your own repository.
git clone https://github.com/YOUR-USERNAME/YOUR-REPOSITORY.git
cd YOUR-REPOSITORY
# Node.js 22.12+; .nvmrc selects Node 22 if you use nvm.
npm ci
npm run dev

Open the local address printed by Astro. For the same checks used by CI and deployment:

npm run build
npm run preview

npm run build includes content validation, Astro type checking, static rendering, link/anchor/asset checks, RSS checks, and search validation. The output directory is dist/.

What is included

• Latest issue, dated archives, canonical paper pages, and configurable research topics.
• Markdown reading notes, mathematics, attributed local figures, and article navigation.
• Browser-side full-text search of notes, issues, research briefs, and enabled resources.
• Draft publication rules and RSS for published issues.
• English or Simplified Chinese interface (language in site.config.json). Content is written in your chosen language; changing the interface does not translate your notes or topic names.
• Optional research resource directory with category navigation, configurable fields, and search.
• CI, deployment instructions, and bilingual prompts for AI-assisted setup and ongoing curation.

The three configuration files

File	What to change
site.config.json	Site title, description, URL, base path, interface language, repository link, search examples, resource visibility
src/data/topics.json	Topic slugs, labels, short labels, and descriptions
src/data/research-profile.json	Research questions, inclusion/exclusion criteria, example papers, preferred sources, language, and selection window

The research profile guides you or your assistant. It does not run a recommender, fetch papers, or configure a scheduler. site.config.json.resourcesEnabled is the actual website switch; the profile's resource preference records editorial intent.

Are databases required?

No. These are two different concepts:

• Research databases/resources are optional things you curate: datasets, corpora, archives, benchmarks, registries, catalogs, or tools. The directory is off by default. Enable it only when relevant to your field; its columns are not tied to genetics or any other discipline. See resources.
• An application database such as PostgreSQL, MySQL, or a hosted database is not used. Markdown and JSON in Git are the source of truth. Adding accounts, private notes, or online editing would be a separate application extension.

Research briefs can still appear in issues and search while the separate directory is disabled.

Add your first paper and issue

npm run new:paper -- my-first-paper
npm run new:issue -- 2026-09-07

The date above is an example; use your intended issue date. Both files start with published: false. Fill in verified metadata and notes, put my-first-paper in the issue's papers list, and review the content. Set both publication flags to true when ready. A paper is visible only when referenced by a published issue. See content guide for complete schemas and examples.

Delete src/content/papers/example-reading-note.md and src/content/daily/2026-01-01.md together when you no longer need the demo. A completely empty catalog is supported.

Documentation

Guide	English	中文
Setup and deployment	Deployment	部署指南
Content and optional resources	Content guide	内容指南
ChatGPT / Codex prompts and workflow	AI workflow	AI 协作流程

No daily paper search, AI generation, or automatic publication schedule is enabled. Git-connected hosts can rebuild when you push content; that is separate from finding new papers.

Commands

Command	Purpose
npm ci	Install locked dependencies
npm run dev	Local development
npm run build	Build and run all standard validation gates
npm run preview	Inspect the built site locally
npm run check:content	Validate content and configuration
npm run check:built	Recheck an existing dist/ build
npm run check:search	Recheck the search index and engine
npm run check:variants	Check subpath hosting, resources, drafts, Chinese UI, and an empty catalog; uses an isolated temporary copy
npm run new:paper -- slug	Create a draft paper without overwriting an existing file
npm run new:issue -- YYYY-MM-DD	Create a draft issue

Attribution and license

Extracted from LucaJiang/genetic-daily-papers at commit e47c6bd5a3f5066f5168e9bdf7b5c9277247c9ae. This repository preserves the reading-site structure and search engine while replacing personal research content, domain-specific resource fields, and deployment identity with neutral configuration and examples. The original repository and its deployment are separate.

Code and original template documentation use the MIT License, retaining the original copyright notice. Third-party papers, figures, and datasets retain their own terms. A draft in a public repository is still public source code: published: false controls the built website, not GitHub access.