Skip to main content

Lifecycle Integration

In one line: Skills chain output-to-input through committed files; subagent-driven development runs one fresh agent per task so context never pollutes across tasks.

The three lifecycle patterns integrate through two mechanisms: the Superpowers skill system and the subagent-driven development model.

Skill Chaining

Each Superpowers skill produces output that serves as input for the next skill in the lifecycle:

SkillOutputConsumed By
/brainstormingProblem analysis, candidate approaches/writing-plans
/writing-plansOne design artifact: specification + ordered task list/test-driven-development or /executing-plans
/test-driven-developmentImplemented code with passing tests/verification-before-completion
/systematic-debuggingRoot cause analysis, reproduction steps/test-driven-development (for the fix)
/verification-before-completionVerified completion evidence/requesting-code-review
/requesting-code-reviewReview findings and dispositions/finishing-a-development-branch or back to implementation
/finishing-a-development-branchMerged branch, updated documentationNext task or session end

The skills do not share state directly. Each skill operates on files (source code, design documents, plan files, test output) that persist in the repository. This file-based integration means that skill transitions are auditable — the design document, the plan, the test output, and the review findings are all committed artifacts that can be inspected after the fact.

Subagent-Driven Development

Subagent-driven development is the default execution model: a fresh subagent per plan task, so context from earlier tasks does not pollute later ones. Three roles (detailed in Section 5.4):

  1. Orchestrator (human or primary session) — reads the plan, dispatches per task, reviews output between tasks, decides proceed-or-revise. Directs and reviews; does not implement.
  2. Implementer (subagent) — receives one task with its context, implements (TDD or code-first), verifies, reports. Builds what it is directed to build.
  3. Reviewer (agent) — checks output for spec compliance and code quality. Validates; does not implement.

This separation of concerns prevents the common failure mode where a single session accumulates bias toward its own implementation and loses the ability to critically evaluate its own output.

When to use /executing-plans instead: Use a single session working sequentially when subagent dispatch is impractical — parallel sessions already running, a plan of 3 or fewer low-complexity tasks, or tightly coupled tasks that repeatedly modify the same code. The choice is pragmatic, not a quality decision: both paths converge at the same verification and review gates.

Lifecycle Transition Triggers:

TriggerLifecycleEntry Point
New feature request or architectural changeFull Cycle/brainstorming
Bug report or test failureBug Fix Cycle/systematic-debugging
Pull request from another developerCode Review CycleStage 1: Spec Compliance
Refactoring initiativeFull Cycle (abbreviated)/writing-plans (skip brainstorm if scope is clear)
Dependency update or migrationFull Cycle/brainstorming (if breaking changes) or /writing-plans (if straightforward)
Documentation updateDirect implementationNo lifecycle required for content-only changes