Skip to main content

Scaffold a complete Vivary agent workspace — the create-t3-app for agent-native workspaces.

Project description

create-vivary

Scaffold a complete Vivary agent workspace: tropo config, strato workspace files, runtime skills, private-memory boundaries, progressive module indexes, and a starter typed graph.

Current release: 0.2.6. Use 0.2.6 for new installs; no migration is expected from 0.2.1, 0.2.2, 0.2.3, or 0.2.5.

Release focus: 0.2.6 adds the knowledge-work preset, capability discovery, optional semantic-memory setup, and doctor memory reporting. Cognee remains an explicit optional policy, not a default dependency or indexing step.

Security hardening: The 0.2.5 line 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.

Install & scaffold

pip install create-vivary==0.2.6              # or: uvx create-vivary@0.2.6 ...
create-vivary my-workspace --preset coding    # interactive wizard on a TTY
create-vivary my-workbench --preset knowledge-work --memory local
create-vivary my-codebase --preset coding --active-context cocoindex-code
create-vivary capabilities --preset second-brain --json
create-vivary doctor my-workspace

# Agent-mode (no prompts, machine-readable output):
create-vivary init . --preset coding --auto --size large --yes --json

# Reconfigure storage on an existing workspace:
create-vivary wizard my-workspace --auto --storage embedded --yes --json

create-vivary <name> is shorthand for create-vivary init <name>; pass init / doctor / wizard / capabilities explicitly whenever you prefer. The same UX is available on npm via the @vivary/create launcher (npm create @vivary@latest my-workspace), versioned in lockstep.

Presets share the same agent OS shell, then seed a different starter graph. Each starter module is generated as modules/<id>/index.md so agents route through a small module index before opening deeper context:

Preset Module First slice Verification
coding codebase local-ci-baseline local-checks
second-brain knowledge-base capture-routine retrieval-smoke
knowledge-work workbench + sources workbench-first-artifact workbench-proof
writing manuscript-system draft-review-loop editorial-review

The command is local-only. With --storage embedded or --auto on a large workspace, it self-installs vivary-tropo[embedded] (LanceDB) with a confirmation prompt, or silently with --yes. Scaffold writes, storage config writes, and stale generated cleanup refuse symlinked destination parents, including when --force is used, so output stays inside the selected target. Use --dry-run to simulate without writing, installing, or cleaning stale files. For scripted storage selection, pass --no-wizard --storage embedded --yes or use --auto; in human mode, the wizard asks and its answers drive storage.

Semantic memory is a separate optional capability. --memory local writes local-only policy and graph nodes; --memory cognee writes Cognee policy and verification docs. Neither option indexes content or sends data anywhere during scaffold, and Cognee is not installed by default. Use create-vivary capabilities --preset <name> --json to show agents the available optional pieces before setup.

For coding workspaces, --active-context cocoindex-code adds an optional CocoIndex-code sidecar profile: active-context skills for Claude/Codex-style agents, local policy docs, graph nodes, and .cocoindex_code/ in .gitignore. It does not auto-install CocoIndex-code, create an index, or enable MCP; the generated docs give the approved ccc init / ccc index path.

doctor validates the generated shell, active privacy ignore rules, module directory indexes, semantic-memory status, and typed graph:

python packages/create-vivary/create_vivary.py doctor sandboxes/coding-demo --json

Developing from source

python packages/create-vivary/create_vivary.py init sandboxes/coding-demo --preset coding
python packages/create-vivary/create_vivary.py doctor sandboxes/coding-demo
python packages/tropo/tropo.py check --root sandboxes/coding-demo

Website & docs: https://vivary.vercel.app/

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

create_vivary-0.2.6.tar.gz (44.5 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

create_vivary-0.2.6-py3-none-any.whl (43.0 kB view details)

Uploaded Python 3

File details

Details for the file create_vivary-0.2.6.tar.gz.

File metadata

  • Download URL: create_vivary-0.2.6.tar.gz
  • Upload date:
  • Size: 44.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.14.3

File hashes

Hashes for create_vivary-0.2.6.tar.gz
Algorithm Hash digest
SHA256 4a40a20f2f0de911731474d107516f2e5c9ddd02ff2f7fd1101eeea5f350305c
MD5 cbdbf5fea629931113f6f71c43a6cc3a
BLAKE2b-256 8bd1caeba9196317740b900ccac5f681605ac5f5e241de1e13c3302457f7c172

See more details on using hashes here.

File details

Details for the file create_vivary-0.2.6-py3-none-any.whl.

File metadata

  • Download URL: create_vivary-0.2.6-py3-none-any.whl
  • Upload date:
  • Size: 43.0 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.14.3

File hashes

Hashes for create_vivary-0.2.6-py3-none-any.whl
Algorithm Hash digest
SHA256 03d564177fe3c96c669fc479149c7b4191ba124909698515e05fb6f54e4f04b9
MD5 62866ab3118e1396d54398262f455388
BLAKE2b-256 aedd546a52381c08b37c73f690b5436ef1639d0b681eac5828f6c408118d3977

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page