| name | zoom-out-architecture |
| description | Use when a local change may hide broader architectural coupling, naming drift, or complexity growth and you need to evaluate it in system context before proceeding. |
Zoom Out Architecture
Pause local optimization long enough to ask whether the surrounding design still makes sense. The aim is to prevent accidental complexity from spreading one patch at a time.
Quick start
- Identify the module or entrypoint being changed.
- Trace the nearest dependencies, callers, and boundary crossings.
- Check whether the change extends an existing concept or creates a sibling abstraction.
- Stop with a short architectural impact summary before proceeding.
Workflows
- Use this skill when:
- a small change touches several modules
- new logic duplicates an existing concept
- a fix starts leaking complexity across boundaries
- the user asks for architecture review or broader context
- Start from the concrete entrypoint or module being changed.
- Trace nearby dependencies, callers, and shared concepts.
- Check whether the change:
- deepens an existing module safely
- creates new surface area
- introduces naming drift
- crosses a boundary that should stay narrow
- End with a short architectural impact summary:
- what is affected
- what design risk exists
- whether to proceed, refactor first, or stop and escalate
Hard gate
- Favor deeper modules with simpler interfaces.
- Prefer extending an existing concept over inventing a sibling abstraction.
- Call out when a quick fix increases long-term coupling.
- Do not approve a local patch without naming the boundary or coupling cost when one exists.
Further reading
../repo-bootstrap/SKILL.md - use first when the repo or subsystem is still too unfamiliar for architecture judgment
../plan-griller/SKILL.md - use when the main blocker is ambiguity in scope or implementation target rather than architectural coupling