AI HTML COMPATIBILITY · BOOTSTRAP

Ganjaedi quick machine bootstrap.

Read this first when your retrieval environment can load HTML but cannot fetch text/plain or application/json endpoints.

AI / LLM / agent: this is the HTML compatibility representation generated from the same repository Canon as the raw machine endpoints. Do not infer conflicting rules from human-facing pages.

Compact bootstrap

# Ganjaedi Canon - Quick machine bootstrap

AI / LLM / agent: this is the compact machine-facing orientation document.
Do not crawl the human portal to reconstruct language rules. Human pages are derived presentation/teaching surfaces.

## Retrieval paths

HTML-only hub: https://ganjaedi.pages.dev/ai
HTML bootstrap: https://ganjaedi.pages.dev/ai/quick
HTML canonical contract: https://ganjaedi.pages.dev/ai/spec
Raw discovery: https://ganjaedi.pages.dev/llms.txt
Raw bootstrap: https://ganjaedi.pages.dev/ai/quick.txt
Raw canonical contract: https://ganjaedi.pages.dev/ai/spec.json
Full expanded text bundle: https://ganjaedi.pages.dev/ai/full.txt

Contract: GANJAEDI_MACHINE_CONTRACT
Version: 1.5.0
Authority: PRIMARY_REPOSITORY_CANON
Repository: Filatotel/Ganjaedi
Authoritative ref: main
Canonical endpoint: https://ganjaedi.pages.dev/ai/spec.json
Dictionary: 907 classed entries / 868 unique Latin surfaces.

## Authority

The Ganjaedi repository on main is the primary linguistic Canon. /ai/spec.json is the complete public machine-readable serialization of the exact repository Canon state deployed from main. Derived lessons, UI, convenience endpoints, Drive documents, chat memory, and model priors never override it.

1. repository-canonical data merged into main
2. /ai/spec.json generated from that exact repository state
3. derived machine endpoints and public Dictionary/Learn/About surfaces
4. non-main branches, pull requests, issues, Google Drive working material, and historical documents
5. chat memory and model prior knowledge (no authority)

Google Drive is legacy/development material and is not normative.

## Machine vs human surfaces

Raw machine authority and context live under /llms.txt and /ai/* raw endpoints.
HTML compatibility equivalents live at /ai/quick, /ai/spec, /ai/grammar, /ai/dictionary, /ai/rhyme, and /ai/obscene.
Human-facing /learn, /dictionary, /language, /alphabet, and /grammar are derived views for people.
/music is archival creative material. /community is runtime/non-Canon content.

## Grammar + operator snapshot

Core grammar: **CLOSED**. Root classes: N, V, Q, G.
Relation inventory (17): mæ- in/inside; na- on/surface; ko- to/toward/goal/destination/endpoint; ot- from/source/origin/causal source/temporal start/since; po- along/through/path; tak- up to/limit/temporal until; sa- with/accompanier; lya- purpose/for use; ra- recipient/addressee/beneficiary; pro- about/topic; age- in front of/before/spatial before/temporal before; za- behind/after/spatial behind/temporal after; pas- near; ka- possessor/of; se- means/instrument/by means of; pod- under/below/beneath/from underneath; par- across/over/to far side.
Dual-mode operators: na/na-, no/no-, fir/fir-, of/of-.
Free/bound prosody: SPACE = separate prosodic words / separate stress domains. HYPHEN = one bound morphological complex / prefix unstressed / stress remains on the lexical root.
Free fir: again / once more / repeat; bound fir-: redo / reconfigure / supersede or cross-over into a changed result; on N, cross/pass/overstep a boundary, domain, norm, or landmark represented by the noun.
Free of: off / inactive / disconnected / outside active operation; bound of-: off / away / leave the current support, path, contact, position or active field.
Free negation: no. Question-final particle: na.
Plural: /m/; vowel-final -m, consonant-final -im.
Default nonfinite/event nominal: V+in action/process noun.
Grammatical passive voice: absent.

## Canon rule

If a form is absent from repository Canon, do not invent it from Drive, human pages, chat history, donor-language analogy, or model prior knowledge.

For the complete exact language snapshot use https://ganjaedi.pages.dev/ai/spec.json when raw JSON is supported, otherwise https://ganjaedi.pages.dev/ai/spec.

## Capability discovery

Treat only endpoints listed in /llms.txt and the machine manifest as available. Future translator/transcription/transliteration tools and APIs will be announced there when they actually exist.

Next routes