Lifecycle Toolkit

In short: AgentSeek standardizes the development workflow around generated apps without owning their runtime.

Context

AI app templates can differ in runtime, frontend, environment variables, and local services.

Developers still need the same basic workflow: create, check, run, inspect, and extend.

Lifecycle map

Stage	Goal	Primary command	Detailed page
Create	Generate an editable app from a maintained template.	agentseek create	Create a Project
Inspect	Understand services, entry points, environment checks, and tasks.	agentseek info	Inspect a Project
Configure	Fill required environment values and install project dependencies.	project tools	Get Started
Check	Verify local readiness before starting services.	agentseek doctor	Check a Project
Run	Start or preview the template-defined local workflow.	agentseek dev	Run Local Development
Extend	Add or adjust project-defined lifecycle tasks.	agentseek task	Run Project Tasks

After agentseek create, most lifecycle commands should run from the generated project root. The generated project owns the .agentseek/lifecycle.toml file that tells AgentSeek which services, checks, and tasks exist.

How it works

AgentSeek provides the command surface. Each generated project provides the lifecycle behavior.

stable command
  -> project lifecycle spec
    -> template-specific behavior

Why it is like this

The command surface stays predictable across templates.

The generated app keeps control of its runtime details. That makes templates free to evolve without adding a new AgentSeek command for every runtime choice.

Consequences for users

• You use the same AgentSeek commands across templates.
• You inspect and change app behavior inside the generated project.
• You use agentseek task when a template exposes extra project tasks.
• You can read the lifecycle spec reference when you need exact field semantics.