curl -s https://www.pathwren.workers.dev/c.json # this page, as JSON
No key, no account, no handshake — every page here has a JSON twin one hop away. Machine doors: 6 keyless GET tools · documents.json · changes · llms.txt · openapi.json · agent card · mcp · a2a
curl -s https://www.pathwren.workers.dev/openapi.json # the document curl -s https://www.pathwren.workers.dev/c/apis-io/openapi.json # the same bytes, tagged apis-io
Every document on this host is also served under /c/<channel>/ — the same bytes, at an address that records which directory or registry sent you. The canonical is always the untagged path.
/c/<channel>/<document> serves /<document>
byte for byte. Not a redirect: a client that does not follow redirects still gets the
document, in one request.Link: <…>; rel="canonical"
naming the untagged address (RFC 6596). Index the canonical; these are
not 241 copies of a page, they are one page with 241 entrances.s|^/c/[^/]+/|/|. That is the whole transform./c/<channel> with no document after it serves the site root.?c=<channel> means the same thing and obeys the
same register. Path form exists because some addresses cannot carry a query string at
all: a Go module path is one, a robots.txt line is another.Measured here, not asserted: One link, 709 fediverse instances — 1,687 client keys from 1,342 operators, counted per instance rather than per impression, after one link was posted to one community. One of five documents about the same 24 hours — the other four are named at the foot of each one — every one of them also at .md and .json, with the figures and the SQL under the data index beside them.
The channel name, plus the same request fields every web server logs: path, status, user-agent, referer, a truncated hash of the address. No cookies, no JavaScript, no fingerprinting, no cross-site anything — the tag is in the URL you were given, which is the only reason it exists. /privacy.html is the full log schema; it is short.
/c is not a Go module root and neither is any path under it.
This host serves no go-import meta tag anywhere, because it
publishes no Go module. A module path of ours ending in .git resolves by the
generic VCS rule to a real repository — see /git — which contains no
.go files and is an archive rather than a library.
GoCommand/1 (+https://go.dev/cmd/go) asked this address
?go-get=1 at 2026-09-05T23:47:40Z while walking the prefixes of
www.pathwren.workers.dev/c/go-module-proxy/git/ai-crawler-index.git,
and got a 404 that meant “no such address” when the true answer was
“a real address, and not a module root”. This page is that answer.
| Tagged address | The document it serves |
|---|---|
https://www.pathwren.workers.dev/c/apis-io/openapi.json | /openapi.json |
https://www.pathwren.workers.dev/c/pypi-registry/openapi.json | /openapi.json |
https://www.pathwren.workers.dev/c/rubygems-registry/openapi.json | /openapi.json |
https://www.pathwren.workers.dev/c/go-module-proxy/openapi.json | /openapi.json |
241 channel names are declared. The register itself is not published — knowing which directory a URL was minted for tells a reader nothing they cannot see from the URL they hold.
This page as markdown: /c.md — the same text, no markup to strip, no JavaScript, no key, CC0. Every page here has one: add .md to any address (also .mdx, <page>.html.md, <page>.html.mdx), or send Accept: text/markdown to this one. All of them in a single index: /sitemap.md.