vac-optimizer/.agents/skills/improve-codebase-architecture/SKILL.md

name, description

name	description
improve-codebase-architecture	Review a codebase for module-deepening opportunities and explain the most useful candidates. Use for an architecture survey or a specified area of architectural friction.

Improve Codebase Architecture

Surface architectural friction and propose deepening opportunities — refactors that turn shallow modules into deep ones. The aim is testability and AI-navigability.

This command is informed by the project's domain model and built on a shared design vocabulary:

• Use /codebase-design to reason about module depth, caller contracts, locality, and justified test seams. Preserve project names such as component, service, or API; explain concepts without imposing a competing vocabulary or adapter count.
• The domain language in CONTEXT.md gives names to good seams; ADRs in docs/adr/ record decisions this command should not re-litigate.

Process

1. Explore

Scope before you scan — YAGNI. Deepening a module pays off by making future changes to it easier, so put extra weight on the parts of the codebase that have recently changed. Decide where to look before you look:

• If the user named a direction — a module, a subsystem, a pain point — take it, and skip the inference below.
• Otherwise, walk back a good stretch of the commit history (git log --oneline) to find the codebase's hot spots — the files and areas that keep coming up — and let those paths pull your attention first. If the changes are scattered with no clear hot spot, widen the net.

Read the project's domain glossary (CONTEXT.md) and any ADRs in the area you're touching first.

Explore directly or delegate bounded independent areas through the available tools when useful. Note concrete friction:

• Where does understanding one concept require bouncing between many small modules?
• Where are modules shallow — interface nearly as complex as the implementation?
• Where have pure functions been extracted just for testability, but the real bugs hide in how they're called (no locality)?
• Where do tightly-coupled modules leak across their seams?
• Which parts of the codebase are untested, or hard to test through their current interface?

Apply the deletion test to anything you suspect is shallow: would deleting it concentrate complexity, or just move it? A "yes, concentrates" is the signal you want.

2. Present the useful candidates

Honor the requested format and output path. A concise Markdown report is enough for a few findings. For a requested visual HTML report, use a fresh temporary path by default and provide its location; open it only when requested.

Use a diagram only when relationships or a before/after change are clearer visually. Prefer static HTML/CSS/SVG or existing renderers; do not add CDN scripts or dependencies for a simple report.

For each candidate, render a card with:

• Files — which files/modules are involved
• Problem — why the current architecture is causing friction
• Solution — plain English description of what would change
• Benefits — explained in terms of locality and leverage, and how tests would improve
• Before / After diagram — only when it materially clarifies the change
• Recommendation strength — one of Strong, Worth exploring, Speculative, rendered as a badge

End the report with a Top recommendation section: which candidate you'd tackle first and why.

Use CONTEXT.md vocabulary for the domain and accurate project names for the implementation. Design terms should clarify the proposed change, not rename existing concepts without need.

ADR conflicts: if a candidate contradicts an existing ADR, only surface it when the friction is real enough to warrant revisiting the ADR. Mark it clearly in the card (e.g. a warning callout: "contradicts ADR-0007 — but worth reopening because…"). Don't list every theoretical refactor an ADR forbids.

For HTML output, read HTML-REPORT.md for a small offline scaffold and relevant diagram options.

Return the requested survey with a recommendation. If the user wants an interactive design session, offer to explore a candidate; otherwise stop at the report.

3. Grilling loop

When the user requests an interview about a candidate, use /grilling for unresolved constraints and decisions. Reuse existing answers and honor a request for a direct design recommendation instead.

When domain-document updates are within scope, use /domain-modeling as decisions crystallize; a read-only architecture report should only recommend these updates:

• Naming a deepened module after a concept not in CONTEXT.md? Add the term to CONTEXT.md. Create the file lazily if it doesn't exist.
• Sharpening a fuzzy term during the conversation? Update CONTEXT.md right there.
• User rejects the candidate with a load-bearing reason? Offer an ADR, framed as: "Want me to record this as an ADR so future architecture reviews don't re-suggest it?" Only offer when the reason would actually be needed by a future explorer to avoid re-suggesting the same thing — skip ephemeral reasons ("not worth it right now") and self-evident ones.
• Want to explore alternative interfaces for the deepened module? Run the /codebase-design skill and use its design-it-twice parallel sub-agent pattern.