nc://webly

Webly: publish a website or share files at a live webly.site link. A Nodecode rewrite of Webly-AI/webly-plugin by Webly-AI.

Get it

Tell the agent "add the webly cell", or run nodecode add nc://webly.

Set it up

Nothing. Publishing without an account works at once: a folder becomes a public page at a *.webly.site address, live for 24 hours and claimable for 7 days. The section is optional:

"webly": { "api_url": "https://api.webly.ai", "timeout": 60, "name": "" }

Use it

The model has three tools:

• webly_publish takes a folder (or one .html file) and answers the live URL. The first publish creates this machine's site; every later one updates the same site at the same URL. Limits: 25 files, 1 MiB each, 5 MiB in all; hidden files and node_modules are left out.
• webly_share takes files or folders and answers a public link for each, plus one folder link, for 24 hours.
• webly_site answers where the saved site stands: its URL, when it goes offline, when it can still be claimed, and what the API says is next.

The same verbs are callable through eval, (help :webly) teaches them, and /webly is the operator's switch:

/webly                     the site's state, with a row that publishes a folder
/webly publish ./site      publish that folder
/webly share ./shots       share a folder as links
/webly claim               open the page that signs in and claims the site

A claim needs a browser: /webly claim prints the page, which carries the site's token in the fragment, so nothing can read it but the page itself. After a claim the site belongs to an account, and publishing again goes through Webly's signed-in tools, which this cell does not carry (see NOTICE).

What is in it

Every request goes to the API the section names and to nothing else, except the presigned upload URLs that API returns. The credential is kept under the home as webly/state.json, mode 0600, and is never printed.

License

Apache-2.0. The helper this rewrites is Copyright (c) Webly: see LICENSE and NOTICE.