Use when a value stream, strategic outcome, portfolio item, or epic-sized change needs to be decomposed into capability increments, business hypotheses, dependencies, architecture concerns, and feature candidates.
Capabilities describe meaningful abilities needed to improve a value stream. They are larger than features, but still need evidence, boundaries, and a route to implementation.
Use superpowers:brainstorming when installed; otherwise use an equivalent design-discovery workflow to explore competing capability options before locking scope.
A good capability:
# Capability: <name>
**Parent value stream:** <name>
**Benefit hypothesis:** If <ability exists>, then <customer/business/operational result> will improve because <reason>.
**Primary users / actors:** <people, systems, teams>
**Capability type:** <business | platform | operational | enabler>
## Scope
- Includes: <abilities and scenarios>
- Excludes: <explicit non-goals>
## Measures
- Outcome: <measure>
- Learning: <what must be validated>
- Guardrail: <quality, security, reliability, cost, compliance>
## Dependencies
- Upstream: <capabilities, systems, decisions>
- Downstream: <teams, systems, rollout constraints>
## Architecture Questions
- <question that may require C4 context/container/component view>
## Feature Candidates
- <feature candidate and reason>
Proceed to shaping-features when installed, or to an equivalent feature-shaping workflow, only when: