Choose by retrieval capability.

Both layers are generated from the same repository Canon. HTML is a compatibility transport, not a separate authority.

HTML-only: /ai → /ai/quick → /ai/spec · Raw-capable: /llms.txt → /ai/quick.txt → /ai/spec.json

HTML compatibility layer.

Use these routes when a browser or retrieval system rejects text/plain, JSON, XML, or other non-HTML responses.

01

Quick HTML bootstrap

Start here when your browser/retrieval tool can read HTML but rejects text/plain or JSON.

02

Canonical contract — HTML

Complete HTML rendering of the same canonical payload published at /ai/spec.json.

03

Grammar — HTML

Operators, grammar, and discourse in an HTML-compatible projection.

04

Dictionary — HTML chunks

Complete dictionary exposed as a small HTML index plus bounded static chunks for retrieval.

05

Rhyme — HTML

Canonical rhyme and lyric-writing guidance.

06

Obscene register — HTML

Canonical obscene-register contract and operator context.

Raw machine layer.

Prefer these endpoints when your environment can retrieve their MIME types directly.

01

Discovery router

Raw text discovery and task-routing index for agents that support text/plain.

02

Quick bootstrap

Compact text/plain machine bootstrap.

03

Canonical specification

Complete authoritative JSON snapshot of deployed repository Canon.

04

Grammar JSON

Structured operators, grammar, and discourse projection.

05

Dictionary JSON

Complete structured dictionary projection.

06

Rhyme JSON

Structured rhyme guidance.

07

Obscene-register JSON

Structured obscene-register projection.

08

Full text bundle

Large expanded machine context.

Route by task.

The HTML route is listed first; the exact raw equivalent is shown beside it.

01

General orientation

HTML: /ai/quick

RAW: /ai/quick.txt

02

Exact generation / translation / validation

HTML: /ai/spec

RAW: /ai/spec.json

03

Lexical lookup

HTML: /ai/dictionary

RAW: /ai/dictionary.json

04

Grammar / operators / discourse

HTML: /ai/grammar

RAW: /ai/grammar.json

05

Rhyme / lyrics

HTML: /ai/rhyme

RAW: /ai/rhyme.json

06

Obscene register

HTML: /ai/obscene

RAW: /ai/obscene.json

Authority and future capabilities.

Repository main remains the primary linguistic Canon. Human Learn, Dictionary and reference pages are derived; Music is archival creative material; Community is runtime content. Future translator, transcription/transliteration, validator and API routes count as available only after they are published in the machine manifest and discovery surfaces.

Current machine contract: GANJAEDI_MACHINE_CONTRACT v1.5.0