Clear content.
Useful access

Find, read, and cite Coolset Academy articles through a small public interface. No account or browser interaction is required.

This is an isolated review website with 214 imported public articles. The live website remains at coolset.com. The import is a snapshot. Each article links to its original source.

Start with the content index

Use this interface to find Academy explanations, compare articles, or retrieve publication dates and original source citations.

Search article titles, descriptions, categories, and authors. Fetch the article body after you find a relevant result.

curl 'https://coolset-website-preview.vercel.app/api/content?q=EUDR&limit=5&offset=0'

Read a published article as Markdown:

curl 'https://coolset-website-preview.vercel.app/api/content/eudr-mixing-and-circumvention-risk-how-to-handle-blended-commodities-cooperatives-and-aggregated-supply-chains/markdown'

Use the returned sourceUrl when citing the live article. Preserve the author, dates, and source links in your answer.

Available routes

GET /api/content

Published article metadata. Supports q, category, limit, and offset.

GET /api/content/{slug}

One published article as JSON, including its sanitized body and original source URL.

GET /api/content/{slug}/markdown

The article as Markdown. HTML blocks preserve tables and other content that needs its structure.

GET /api/markdown?path=/

A public page as Markdown. The same route serves negotiated website requests.

GET /openapi.json

The OpenAPI contract for these public operations.

GET /.well-known/api-catalog

A catalog that links the interface, contract, and this guide.

GET /llms.txt

A short guide to the website and its public content.

GET /llms-full.txt

All imported public Academy articles in one Markdown response. Use individual articles for smaller requests.

Predictable reads

The index returns total, offset, limit, and items. The default page contains up to 20 articles.
Use a limit from 1 through 100 and an offset of zero or greater. Increase the offset by the number of returned items.
Search uses the first 200 characters and matches metadata, not the full article body. A search with no matches returns an empty list.
Invalid pagination returns status 400 and a JSON error message.
An unknown or unpublished article returns status 404 and a JSON error message.
Relative links in the response resolve against this preview domain.
The explicit Markdown route has its own URL. Public website pages also accept Accept: text/markdown. HTML remains the default.
Negotiated page responses include Vary: Accept. A missing page returns status 404 with recovery links in Markdown.

This prototype has no documented request quota. Keep requests bounded. If the hosting service returns 429, respect its retry instructions.

The contract can change during development. Production versioning and change notices remain a release task.

What this interface supports

Read public content without authentication. Retrieve facts and source links for a human to assess.

These routes do not edit articles, publish content, submit forms, or book meetings. They do not expose the production application's services.

Article publication dates do not confirm that a regulatory statement remains current. Check the cited sources before relying on time-sensitive information.

Our verification target

The first scan scored 74/100 on 15 September 2026. We target 100/100 and successful real retrieval tasks. Later fixes need a new scan.

This preview uses noindex to prevent search indexing. A public scan can observe that instruction and report a different result from production.

A benchmark score is one signal. We also check whether agents find the right article, preserve its sources, and handle missing content correctly.

Read the scan report, the methodology, or the project roadmap.