@archstone/init

Capability inference from an existing API — reads an OpenAPI document, asks the questions no spec can answer, and drafts a Capability Definition Language (CDL) manifest that the real compiler has already compiled. It writes nothing at all if that manifest does not compile.

Part of Archstone, an open-source Capability Platform: a company describes what it can do in CDL (business only, no integration code); Archstone compiles that to a target-agnostic IR; an emitter turns the IR into tools an AI agent can discover and call. This package is the onboarding path into that pipeline — most users should install @archstone/cli and run archstone init, rather than depending on this package directly.

Usage

npm install -g @archstone/cli
archstone init path/to/openapi.yaml --out manifest --company acme --domain catalog

The one answer it never guesses is effect — read, write or irreversible is the difference between looking up a price and charging a card, and no spec says which. Where a response could honestly be read two ways (which array in a payload is the actual result, and which is diagnostics?), it asks rather than picking. With --probe it will also make one read-only call to your real backend and record a genuine fixture, so archstone verify has something true to replay later.

See the main repository README for the full CDL format, worked examples, and a GIF of init running end to end against a demo spec.

Package structure

The root export (@archstone/init) is pure — no node:fs, no HTTP, no terminal, no clock. It carries the draft model and the OpenAPI adapter, and derives every fact mechanically from the source or a human answer, never from an LLM: the same input produces the same output on every run. @archstone/init/loop is the only entry point that touches the filesystem — it owns the closed compile loop that validates a draft before anything is written.

License

Apache-2.0