Seeds

Predefined starting states for digital twins.

A seed is a predefined initial state for a twin — repos, issues, customers, channels, whatever your test needs. Loading a named seed beats hand-rolling state with archal twin start and twenty create_* calls.

Three ways to set state: a named built-in seed, plain English in a scenario’s ## Setup (Archal generates the seed), or a custom JSON file.

Using seeds

In a scenario:

## Config
twins: github
seed: small-project

Or describe state in ## Setup:

## Setup
A GitHub repository called "acme/webapp" with 20 open issues.
8 have not been updated in 90 days. 4 of those have the label "keep-open".

From the CLI:

archal run scenario.md --seed small-project

In a twin session:

archal twin seed github small-project
archal twin seed github --file ./custom-seed.json

Built-in seeds

Each twin ships seeds for common scenarios.

GitHub

Seed	Description
empty	Clean slate
small-project	One repo with a few issues and PRs
enterprise-repo	Org with teams, CODEOWNERS, branch protection
stale-issues	Repo with issues of varying ages
large-backlog	Many issues for triage testing
merge-conflict	PR with merge conflicts
ci-cd-pipeline	Repo with GitHub Actions workflows

Slack

Seed	Description
empty	Clean workspace
engineering-team	Engineering channels, threads, users
busy-workspace	Many channels and threads
incident-active	Active incident channel with ongoing thread

Stripe

Seed	Description
empty	No customers or products
small-business	A few customers, products, active subscriptions
checkout-flow	Products and prices set up for checkout
subscription-heavy	Many subscriptions across lifecycle states
subscription-lifecycle	Trial, active, past-due, cancelled

Jira

Seed	Description
empty	Clean project
small-project	One project with a board and issues
enterprise	Multiple projects, sprints, epics, custom fields
sprint-active	Active sprint with issues in various states
large-backlog	Many issues for backlog grooming

Linear

Seed	Description
empty	Clean workspace
small-team	One team with a few issues
engineering-org	Multiple teams, projects, cycles
multi-team	Cross-team project setup
busy-backlog	Large backlog for triage

Supabase

SQL files (not JSON). They create tables and insert data directly.

Seed	Description
empty	Empty database
small-project	Basic tables and data
saas-starter	Users, plans, subscription tables
ecommerce	Products, orders, customers
edge-cases	Schema with tricky edge cases
vaultline-ecommerce	Full e-commerce schema for security testing
rls-bypass-migration	RLS policy testing
bulk-user-deletion	User deletion workflow data
fired-employee-access	Access control after termination
feature-flag-override-mismatch	Feature flag edge cases
migration-poisoned-comment	Migration with hidden RLS bypass

Google Workspace

Seed	Description
empty	Clean account
assistant-baseline	Inbox with a few emails and some calendar events
gmail-busy-inbox	Many emails across categories
calendar-packed-week	Busy calendar
draft-send-cleanup	Drafts and sent items

Ramp

Seed	Description
empty	Clean account
default	Basic card and transaction setup

Scenario-specific seeds (security, cross-service) load automatically when you run their matching scenarios.

Custom seeds

archal twin seed github --file ./my-seed.json

JSON structure matches the twin’s state schema: each top-level key is an entity collection, value is an array of objects.

Related

• Twins overview
• Twin sessions
• Writing scenarios