Telisik Try it live
Web · Installable (PWA)live

Search the sources. Nothing generated, nothing sponsored.

A search page for questions where the source matters. One query goes from your browser to up to 23 public libraries of human-written material. Results are ranked with a formula printed on the page and quoted word for word.

Web · installable · no account
telisik · live demolive demo
Telisik screenshot
2platforms: Web, Installable (PWA)
MITopen source licence
Why it exists

Every result shows its working.

Telisik (Indonesian for "to look into closely") is a search page for questions where the source matters.

The results are pooled, ranked with a formula printed on the page, and shown with the exact passage, the date and the reason each one sits where it does.

What's inside

What you get.

23 sources, one query

Ten every time: Wikipedia, Stack Exchange, Hacker News, GitHub (GitLab as fallback), OpenAlex (Crossref as fallback), Europe PMC, Internet Archive, Open Library, the Mwmbl web index, Wikibooks. Thirteen more when the query or lens calls for them: Wiktionary, Wikinews, Wikiquote, Wikivoyage, Wikisource, Zenodo, ClinicalTrials.gov, GOV.UK, the Federal Register, the open-data catalogues of the European Union, Canada and Australia, and full-text search inside scanned books.

In the sources' own words

Up to three sentences copied character for character from the top results, each numbered to the result it came from.

Why ranked

Every result says in one line why it is there and opens the arithmetic behind its position: text match, evidence the source reports, the source's own order.

Before AI

One switch limits results to items dated up to 30 November 2022. It is a date filter, not a detector.

Lenses

All, Learn, Research, Build, Discuss, Official choose and re-weight the sources for the job.

Saved copies

If a source fails or runs out of quota, the last answer this browser got for the same search is shown, labelled with its age.

Did you mean

Wikipedia's spelling suggestion is passed on when it differs from what was typed.

Cite

A ready citation for any result, one press to copy.

Operators and bangs

"exact phrase", -word, in:papers, in:gov, in:dict, se:physics, before:2020, after:2019-06. Bangs such as !g, !ddg, !maps and !yt hand the query to another engine, so it can be your default search.

Works offline as a shell

Installable; the page opens with no connection and says which sources did not answer.

Run it yourself

Run it yourself.

From the README’s “Run it yourself” section. Full options, requirements and troubleshooting are there too.

Open the README
shell
$ git clone https://github.com/juw12/telisik
$ cd telisik
$ python -m http.server 8080      # then open http://localhost:8080/
Your data

Private by construction

  • No server, no account, no key, no cookies, no analytics.
  • No remote scripts or fonts, no inline script, and a content security policy that lists exactly the source hosts.
Good to know

Honest limits.

Straight from the README, so you know what you're getting before you try it.

  • It is not a crawl of the whole web.
  • It is not a news wire. The only news source is Wikinews.
  • The government sources cover a few places.
  • It matches words, not meaning.
  • It sees titles and short excerpts, not full pages.
Under the hood

How it's built.

Built with

Language
JavaScriptHTML/CSSPython
Platform
PWAOffline service worker
Testing
Automated tests

74 files in the repo · last updated 14 Nov 2026.

Browse the code

Recently shipped

  1. Twenty sources, measured ranking, saved copies, quota notes; fix Hacker News dedupe

  2. Sumber: sourced search, nothing generated

Full history on GitHub
Questions

Good questions.

What does it run on?

Web, Installable (PWA). You can try it in the browser with nothing to install.

How do I get it running?

The commands in the “Run it yourself” box above come straight from the README. Paste them into a terminal; the README covers options and troubleshooting.

Where does my data go?

No server, no account, no key, no cookies, no analytics. No remote scripts or fonts, no inline script, and a content security policy that lists exactly the source hosts.

What doesn't it do yet?

It is not a crawl of the whole web. It is not a news wire. The only news source is Wikinews.

Is it finished?

Not yet: it's marked “live” and still changing. The README always has the current state.

Is it free?

The source is public on GitHub under the MIT licence.

Who made it?

Juwen Lie, an R&D mechanical engineer in Batam, Indonesia, as an independent personal project.

Keep exploring

More from the same workbench.

Search the sources.

Search the sources. Nothing generated, nothing sponsored.