09.06 — Public resources#

Full table of public URLs exposed by the Holy Book. For humans, LLMs, scrapers.

Listing (potentially non-exhaustive)#

llms.txt (emerging standard)#

Spec: https://llmstxt.org/ (proposed in 2024).

Typical format:

# The Ocarina Holy Book

> Ocarina is Igor's automated web browser testing framework.

## Pages

- [What is Ocarina?](https://mojo-molotov.github.io/ocarina-holy-book/what-is-it) — Test frameworks all made the same bad bet. Ocarina does the opposite.
- [First feedbacks](https://mojo-molotov.github.io/ocarina-holy-book/first-feedbacks) — The first wave of criticism Ocarina ever received.
- [First steps](https://mojo-molotov.github.io/ocarina-holy-book/setup) — Onboarding tutorial.
- ... (every page)

## Resources

- [llms-full.txt](https://mojo-molotov.github.io/ocarina-holy-book/llms-full.txt) — All content in plaintext.
- [CLAUDE.md](https://mojo-molotov.github.io/ocarina-holy-book/CLAUDE.md) — AI rules.
- ... etc

An LLM visiting the site can read llms.txt first and know everything that exists.

llms-full.txt#

The plain text content of every page, concatenated. Lets an LLM (Claude, ChatGPT, etc.) fetch the docs in a single HTTP fetch rather than N fetches.

Typical volume: ~500 KB of plain text (for ~5,000 Markdown lines).

robots.txt#

# docs/public/robots.txt
User-agent: *
Allow: /

Everything is crawlable. No restriction on bots or LLMs. “If you want to know, read, RTFM.” No captcha, no paywall.

.nojekyll#

# docs/public/.nojekyll

(Empty file.)

GitHub Pages uses Jekyll by default. Jekyll was converting files into HTML in a way that broke the skills. This disables it.

.spa#

# docs/public/.spa

VitePress artifact with no special interest, implementation detail.

favicon.ico and logo.png#

The favicon.ico shows in the browser tab. The logo.png appears in the banner, top-left (@sugarat/theme).

LLM flow#

1. GET /llms.txt                              → resource list
2. (option A) GET /llms-full.txt              → everything in one fetch
   (option B) GET /CLAUDE.md                  → compact rules
   (option C) GET /<page>                     → specific page
3. If question on internal mechanics:         → fetch Ocarina's source on GitHub
4. If question on usage:                      → fetch ocarina-example
5. If question on AI usage:                   → fetch ocarina-with-ai-example

Holy Book reminder#

The Holy Book (chapter “Using Ocarina with AI”) lists the URLs in cleartext at the bottom:

Exposed resources#

• https://mojo-molotov.github.io/ocarina-holy-book/llms.txt
• https://mojo-molotov.github.io/ocarina-holy-book/llms-full.txt
• https://mojo-molotov.github.io/ocarina-holy-book/CLAUDE.md
• https://mojo-molotov.github.io/ocarina-holy-book/CLAUDE.slim.md
• https://mojo-molotov.github.io/ocarina-holy-book/ocarina-ru.pdf
• https://mojo-molotov.github.io/ocarina-holy-book/ocarina-en.pdf
• https://mojo-molotov.github.io/ocarina-holy-book/ocarina-fr.pdf

A public invitation to LLMs: “Here is what we expose for you, help yourselves.”

Consistency with the political stance#

See ../11-independence/

The author wants AI to consume the docs. A deliberate choice:

• Rather than resisting LLM indexing (paywall, captcha, DRM), we invite.
• Rather than obscuring project rules ("we don’t share our CLAUDE.md"), we publish.
• Rather than gatekeeping ("sign up to access"), we open everything (robots: Allow: /).

The embodiment of Code is Law and sovereign auditability.