engleash 0.2.2

Raw English Native Programming — English is the source code.

English / REN

The user speaks intent in English; REN compiles it into lawful IR; CI/CD executes verified changes; receipts preserve memory; and the system evolves only through human-approved pull requests.

Core Invariant

No English changes reality until it becomes typed IR, passes law, produces tests, writes receipts, and survives human-governed CI/CD.

Seven Organs

1. REN Mouth            — normal English input
2. REN Compiler         — English → intent graph → typed IR
3. REN Law              — policy, type, effect, permission, privacy checks
4. REN Backend Matrix   — IR → Python, TypeScript, Swift, Rust, Go, C/C++, Bash, SQL, YAML, Docker, Terraform, WASM, GitHub Actions, GlyphVM, .aeglyph
5. REN CI/CD Runtime    — governed execution substrate
6. REN Memory           — receipts for every action and artifact
7. REN Evolution Loop   — failures → issues → GAG repair branches → PRs → human merge

Directory

english/
  ren/
    specs/        — type, effect, policy, security model
    schemas/      — intent graph, effect, receipt, policy decision
    compiler/     — parser, IR builder, validator, backend targets
    lexicon/      — word/phrase endpoint registry, Oxford-style adapter
    stdlib/       — reusable English command libraries
    runtime/      — issue awakener, approval gate, rollback manager
    security/     — threat model, sandbox rules, forbidden effects
  programs/       — .ren English programs
  tests/          — compiler + shell + lexicon tests
  receipts/       — generated proof artifacts
  .github/workflows/ — compile, policy, test, self-heal
  aeglyph_shell/  — safe glyph terminal + model-free render runtime
  display_substrate.md — display-native computation thesis

Quick Test

python3 english/tests/test_ren_compiler.py
python3 english/tests/test_aeglyph_shell.py
python3 ren/compiler/ren_compiler.py ren/programs/rm_recovery.ren ren/receipts --api

Speech Agent

Install with speech dependencies:

python3 -m pip install english-language[speech]

Speak and execute:

english listen

Or in Python:

import english
result = english.speech_to_english()
print(result["transcript"])
print(result["result"])

Cross-Language Usage

Install once:

python3 -m pip install english-language

Python

import english
english.run_english("Check availability every 4 hours.")
english.lookup_word("keep")

CLI (any language)

english run "Check availability every 4 hours."
english lookup keep
english compile "1. Check availability every 4 hours."
english stats

Node.js

const english = require('./bindings/node/english.js');
english.runEnglish("Check availability every 4 hours.");
english.lookupWord("keep");

REST API

curl http://localhost:7277/api/lexicon/word/keep
curl http://localhost:7277/api/lexicon/phrase/keep+alive
curl http://localhost:7277/api/lexicon/word/run/senses
curl -X POST http://localhost:7277/api/ren/compile -H 'Content-Type: application/json' -d '{"text":"1. Check availability every 4 hours."}'

C / Go / Rust / etc.

Call the english CLI as a subprocess and parse JSON. The CLI is the stable cross-language interface.

FILE* f = popen("english run 'Check availability every 4 hours.'", "r");
// read JSON from f

Final Sentence

English is source. REN IR is law. CI/CD is runtime. Receipts are memory. Every programming language is a backend. Human-approved PRs are evolution.