@i-language/spec

I-Lang Protocol Specification v3.0 Final -- The native language of artificial intelligence.

88 verbs, 29 modifiers, 14 entities, 13 Greek aliases. Two syntaxes: [] for operations, :: for declarations. Zero ambiguity. Its parser is the LLM itself.

Install

npm install @i-language/spec

Usage

const ilang = require('@i-language/spec');

// Look up a verb
ilang.getVerb('READ');
// { verb: 'READ', alias: null, input: 'null/str/map', output: 'str/bytes/list/map', sideEffect: false, meaning: 'Read content from source', category: 'data_io' }

// Resolve Greek alias
ilang.resolveAlias('φ');  // 'FILT'
ilang.resolveAlias('λ');  // 'MAP'

// Check if something is a valid verb
ilang.isVerb('SCAN');     // true
ilang.isVerb('FAKE');     // false

// Get all verbs in a category
ilang.getVerbsByCategory('transform');
// [{ verb: 'FMT', ... }, { verb: 'CONV', ... }, ...]

// All 8 categories
ilang.categories;
// ['data_io', 'transform', 'analysis', 'generation', 'control', 'output', 'meta', 'system']

// Raw data
ilang.verbs;      // all 88 verbs
ilang.modifiers;  // all 29 modifiers
ilang.entities;   // all 14 entities
ilang.aliases;    // all 13 Greek aliases

ESM

import { getVerb, verbs, resolveAlias } from '@i-language/spec';

Direct JSON import

const verbs = require('@i-language/spec/verbs');
const modifiers = require('@i-language/spec/modifiers');

What is I-Lang?

I-Lang is the native language of artificial intelligence. It reduces semantic loss between human intent and machine execution. Built from symbols already inside every LLM's training data.

88 verbs across 8 categories: Data I/O, Transform, Analysis, Generation, Control, Output, Meta, System
Two syntaxes: [VERB:@ENTITY|mod=val] for operations, ::DECLARATION{} for state
13 Greek aliases: φ = FILT, λ = MAP, Σ = MERGE, Ω = OUT, etc.
Zero ambiguity: Every instruction has exactly one interpretation
Works on ChatGPT, Claude, Gemini, DeepSeek, and any LLM

Full specification: SPEC.md

Website: ilang.ai

License

MIT