Files
zeekay 1943f1c093 Hanzo Commerce
Multi-tenant commerce and billing: orders, subscriptions, invoices, plans,
usage metering and a credit ledger. One Go binary. MIT OR Apache-2.0.

Claude-Session: https://claude.ai/code/session_01SAtaDFm8VDiWrhYHNNtxMw
2026-09-23 21:52:16 -07:00

35 lines
1.3 KiB
Plaintext

.git
.github
.gitignore
.dockerignore
*.md
LICENSE
# Installed dependencies and build output are things the image PRODUCES, never
# things it consumes. At every depth — `node_modules` alone matches only the
# root, and `app/*/node_modules` only the workspace packages, which left
# `app/node_modules` (the pnpm workspace root, where .pnpm and .modules.yaml
# actually live) copied into every build.
#
# That was not merely wasteful, it was silent corruption. The hanzo.yml
# `admin-typecheck` gate runs `cd app && pnpm install` ON THE RUNNER, and the
# Test step runs before the Build step, so by the time the context was packed
# the runner had already created app/node_modules. COPY carried the workspace
# store in while excluding each package's link farm, and pnpm inside the image
# found a modules dir it could not reconcile:
#
# Scope: all 14 workspace projects
# ? The modules directories will be removed and reinstalled from scratch.
# Proceed? (Y/n)
# DONE 4.1s
#
# No TTY, so the prompt read EOF and pnpm exited ZERO having installed
# nothing. `next` was then not on PATH and the admin build died with
# "Local package.json exists, but node_modules missing". The install that
# reported success is the bug; the context is where it came from. 1.44GB of
# it, measured on the wire.
**/node_modules
**/.next
**/.turbo
deploy/
sdk/