HTML Docs Publishing Service

A tiny public home for self-contained agent-created plans, reports, diagrams, and implementation guides.

Quick start

html-docs publish ./report.html
html-docs list
html-docs search "migration plan"
html-docs get report ./downloaded-report.html
html-docs remove report

The file name becomes the public slug unless a slug is supplied explicitly:

html-docs publish ./output.html quarterly-platform-review

Commands

Command	Purpose
`publish FILE [SLUG]`	Create a document or atomically replace one at the same slug.
`update FILE [SLUG]`	Replace an existing document; fail if it does not exist.
`remove SLUG`	Delete a document and refresh the catalog.
`list [--json]`	List titles, slugs, timestamps, sizes, and public URLs.
`search QUERY [--json]`	Search titles, slugs, and visible HTML text.
`get SLUG [OUTPUT]`	Retrieve a public document to stdout or a local file.
`url SLUG`	Print the shareable URL without fetching it.

Publishing contract

Publish one self-contained .html file. Inline its CSS, scripts, diagrams, and small image assets whenever practical so the document remains portable and dependable.

Everything published here is intentionally public. Never include credentials, private customer data, tokens, internal-only URLs, or other secrets.

How writes are protected

Reading is public. Publishing, updating, and removal use the Oracle server's existing SSH key authentication. Files are written atomically, and the public catalog plus JSON manifest are rebuilt after every mutation.