Site ↗
Documentation sections
About the project · 0.8.1

How the Admingen website uses content

How the landing/docs use a JSON snapshot, where articles come from and when the website needs updating.

On this page

Admingen is a Headless CMS generator: it creates a Go backend, React admin and REST API from a project model. A client website independently chooses live external API access or a prepared data snapshot.

The new landing and documentation in admingen-landing use a JSON snapshot in the website repository. This site-specific decision does not limit generated applications. Admin tokens and database credentials are not copied into the website.

Documentation snapshot contents

Articles come from local CMS documentation records: DocSection defines sections and DocArticle defines articles. The snapshot includes section, slug, title, summary, Markdown, order, language and contract version. Existing slugs retain recognizable article links.

There are six sections: Start, UI interface, Concepts, CLI interface, Operations, About the project. Public snapshots contain published articles only. Users, authentication data, secrets and other application tables are not exported.

JSON is display content. It changes neither the CMS model, running database nor API rules.

Updating the website

  1. Edit the DocArticle or corresponding project documentation source.
  2. Keep its slug if the address should remain unchanged.
  3. Prepare a current documentation JSON snapshot and review its contents.
  4. Update the website through its regular build/release process.

Saving a CMS record does not update JSON already released with the website. That requires a new snapshot and website release. Editing record text requires no CMS Generate; changing CMS structure or API rules requires Generate and the corresponding backend release.

Article pages and navigation

The catalog is built from sections and article metadata. Open articles display Markdown; headings generate the table of contents. JSON-version search and navigation use the saved snapshot.

/docs/<slug> links identify documentation articles. For a separate documentation domain, configure client addresses and direct route handling. JSON does not create DNS, servers or certificates.

Markdown contains screenshot:<ключ> markers. These designate illustration slots, not image URLs. Map them to real images or show labelled screenshot placeholders; do not substitute invented product screens.

If your website needs a live CMS API

Generated external APIs remain a supported content-delivery option. Enable required operations and mandatory read conditions in Generator UI → entity settings → Access. For example, status = published restricts results to published records. Virtual entities and nested collections have their own rules; source rules are not automatically inherited.

Public Nginx caching, if explicitly enabled, applies to anonymous external List/Get. Writes do not automatically clear it, so old responses may remain until TTL expiry. This does not update a static JSON snapshot: only changing the file and releasing the website changes its contents.

For a live client, use the API contract. Access is described in Content / Access; caching and connectivity in API performance.