ultrerp-init

name	UltrERP-init
description	Use when the operator says "/UltrERP-init", "init the ERP", "bootstrap the ERP", "first time setup", "fresh install", or "set up from scratch". This skill bootstraps a freshly cloned UltrERP repo — installing dependencies, creating the database, generating the schema, and creating the first admin user. For ongoing operator tasks after init, use /UltrERP-ops. For migration management, use /UltrERP-migrate. For legacy data import, use /legacy-import.
argument-hint	[check|prereqs|install|init-db|generate-migrations|full] [--env <path>]

Use this skill when bootstrapping a freshly cloned UltrERP repository from scratch.

This skill is idempotent and safe to re-run. It detects already-installed components and skips them.

What this skill does (in order)

1. Check prereqs — PostgreSQL running, required tools installed
2. Install dependencies — frontend (pnpm install) and backend (uv sync)
3. Create .env — scaffold from .env.example if missing
4. Configure company — prompt for name, country, currency, fiscal year; write COMPANY_* vars to .env (consumed by backend's company seed on first boot)
5. Create database — CREATE DATABASE ultr_erp if it doesn't exist
6. Generate migrations — alembic revision --autogenerate from SQLAlchemy models
7. Apply migrations — alembic upgrade head
8. Seed settings — default app settings
9. Create admin user — prompt for email, display name, password

Operating rules

1. Detection is always free. Any check-* or prereqs step reads state only. Run without confirmation.
2. Write actions require confirmation for database creation, migration generation, admin user creation. Show exactly what will happen and wait for "yes" or "run it".
3. Follow the order. Skipping steps causes failures downstream.
4. Idempotent. If a step is already done, report "already done" and skip.
5. Fail with remediation. Report the error and suggest how to fix it.

Confirmation checklist before any write action

Show:

• What will be done
• The exact command that will be run
• What success looks like
• What failure could look like and how to recover

Then wait for explicit confirmation.

Sub-commands

Sub-command	Writes?	Description
prereqs	No	Check PostgreSQL, pnpm, uv, Python versions
install	Yes	Install frontend and backend dependencies
init-db	Yes	Create the ultr_erp database if it doesn't exist
generate-migrations	Yes	Run alembic revision --autogenerate to create schema from models
full	Yes	Run the complete init sequence (prereqs → install → init-db → migrate → seed → admin)

Execution flow: full

1. Run prereqs — report what is/isn't installed
2. Confirm and run install — pnpm install + uv sync
3. Confirm and run configure-company — prompt for company details, persist to .env
4. Confirm and run init-db — create database
5. Confirm and run generate-migrations — create initial schema migration
6. Confirm and run apply-migrations — alembic upgrade head
7. Confirm and run seed-settings — seed default app settings (the backend lifespan will seed the company from COMPANY_* env vars on first server start)
8. Confirm and run create-admin — prompt for email, display name, password, then insert user
9. Offer check-health via /UltrERP-ops to verify
10. Offer /legacy-import if legacy staging data will be imported

Company configuration prompt

Run after .env is in place, before init-db. Skip if a company already exists in the DB (re-running on an initialized install).

Prompt the operator for:

Field	Env var	Default	Notes
Company name	COMPANY_NAME	My Company	Shown on invoices; do not ship the default to production
Country	COMPANY_COUNTRY	Taiwan	Free-text, max 100 chars
Default currency	COMPANY_DEFAULT_CURRENCY	TWD	ISO 4217 (3 uppercase letters)
Fiscal year start month	COMPANY_FISCAL_YEAR_START	1	Integer 1–12

Validate before writing:

• name: non-empty, ≤200 chars
• currency: matches ^[A-Z]{3}$
• fiscal year: integer in [1, 12]

Write the four COMPANY_* vars to .env (update in place if already present, append otherwise). The backend reads them on first startup via seed_company(); once a company row exists, subsequent edits to these vars are ignored — the operator must update through the Settings UI / PATCH /api/v1/settings/company.

Prereqs check commands

# PostgreSQL
pg_isready -h localhost -p 5432 2>&1 || echo "PostgreSQL not running"

# pnpm
pnpm --version 2>&1

# uv
uv --version 2>&1

# Python
python3 --version 2>&1

# uv sync (check backend deps)
cd backend && uv sync --dry-run 2>&1

# .env exists
ls .env.example 2>&1 && echo "env example exists"

Database creation command

psql postgresql://ultr_erp@localhost:5432/postgres -c "SELECT 1 FROM pg_database WHERE datname = 'ultr_erp'" 2>&1
# If row not found:
psql postgresql://ultr_erp@localhost:5432/postgres -c "CREATE DATABASE ultr_erp" 2>&1

Migration generation command

cd backend && uv run alembic -c ../migrations/alembic.ini revision --autogenerate -m "initial_schema"

Migration apply command

cd backend && uv run alembic -c ../migrations/alembic.ini upgrade head

Invoking other skills

• After full completes, use /UltrERP-ops for health checks.
• After full completes, offer /legacy-import for historical data.
• For migration management (upgrade, downgrade, stamp, resolve heads), use /UltrERP-migrate.

Use this skill only for first-time bootstrap. For ongoing operations use /UltrERP-ops.