@vivary/create
v0.2.5
Published
Scaffold a complete Vivary agent workspace — the create-t3-app for agent-native workspaces. Runs the Python create-vivary via uv/pipx.
Maintainers
Readme
@vivary/create
The create-t3-app for agent-native workspaces. Scaffold a complete Vivary
workspace — typed knowledge graph (tropo), agent OS (strato), and starter graph — in
one command. Generated modules use modules/<id>/index.md routers so agents load
context progressively.
Current release: 0.2.5. Use 0.2.5 for new installs; no migration is expected from 0.2.1, 0.2.2, or 0.2.3.
Security hardening: 0.2.5 validates active .gitignore rules for USER.md,
MEMORY.md, memory/*, and heartbeat-reports/*; scaffolds private heartbeat report
storage; and refuses symlinked or out-of-workspace scaffold, storage, and cleanup
paths.
npm create @vivary@latest my-workspace -- --preset coding
npm create @vivary@latest my-codebase -- --preset coding --active-context cocoindex-code
# or
npx @vivary/create@latest my-workspace --preset coding
# Agent-mode (no prompts, machine-readable output):
npx @vivary/create@latest init . --preset coding --auto --size large --yes --json
# Reconfigure storage on an existing workspace:
npx @vivary/create@latest wizard my-workspace --auto --storage embedded --yes --jsonPresets: coding · second-brain · writing.
A bare npm create @vivary@latest <name> maps to the init subcommand; you can also
pass init / doctor / wizard explicitly (e.g. npm create @vivary@latest doctor my-workspace).
On a terminal that supports input, init runs a short wizard to pick a storage tier.
For scripted storage selection, pass --no-wizard --storage embedded --yes or use
--auto; in human mode, the wizard asks and its answers drive storage.
How it works
This package is a thin launcher: it runs the Python create-vivary scaffolder via
uv (uvx) or pipx, so the
scaffolder stays one source of truth in Python while you get a Node-native entry
point. Python 3.11+ and uv (or pipx) must already be installed.
Prefer Python directly? uvx [email protected] my-workspace --preset coding — a bare
target defaults to init there too (the PyPI create-vivary is versioned in lockstep
with this launcher) — or pip install create-vivary==0.2.5.
For coding workspaces, --active-context cocoindex-code scaffolds optional
CocoIndex-code guidance and ignored sidecar state. It does not auto-install, index, or
enable MCP; the generated docs give the approved ccc init / ccc index path.
License
MIT.
Website & docs: https://vivary.vercel.app/
