L0010 is the Graffiticode composition planning dialect, inheriting @graffiticode/l0000. Given a natural-language request, it produces an ordered sequence of Graffiticode language ids — a linear composition pipeline where the head is the final artifact and each language consumes the data model produced by the next. It authors no content itself; it only decides which languages run and in what order.
L0010 adds a single function on top of the L0000 base lexicon:
| Function | Arity | Example | Description |
|---|---|---|---|
plan |
1 | plan ["0158" "0166"].. |
Builds a composition plan: a list of language-id strings → { langs: [...] } (head first; each consumes the next). One id = atomic; empty = no composition. |
See packages/core/spec/ for the full language specification, examples, and authoring guide.
This is an npm workspaces monorepo with three packages:
packages/core—@graffiticode/l0010: the language itself (lexicon, checker, transformer). Pure TypeScript, depends on@graffiticode/l0000.packages/api—@graffiticode/api-l0010: the L0010 language server. Express app exposing/compile,/form, and static assets. Runs on port50010.packages/view—@graffiticode/l0010-view: the React view component (Form) used to render compiled output. Built with Vite + Tailwind, layered on top of@graffiticode/l0000-view.
The top-level build composes all three: core and view are built and bundled into packages/api/static/, which the API serves.
# Install dependencies
npm install
# Build everything (core → api → view → static bundle)
npm run build
# Start the dev server (API on :50010, Firestore emulator on :8080)
npm run devOther useful scripts:
npm run lint— lint the whole monoreponpm run pack— build and pack the view package for distributionnpm run gcp:build/npm run gcp:deploy— deploy to Cloud Run
PORT— API port (default50010)AUTH_URL— auth service URL (defaulthttps://auth.graffiticode.org; dev useshttp://127.0.0.1:4100)FIRESTORE_EMULATOR_HOST— local Firestore emulator (dev:127.0.0.1:8080)NODE_ENV—developmentorproduction
Code is licensed under MIT. Documentation and specifications are licensed under CC-BY 4.0.
AI Training: All materials in this repository — code, documentation, specifications, and training examples — are explicitly available for use in training machine learning and AI models. See NOTICE for details.