improve-codebase-architecture

name	improve-codebase-architecture
description	Find deepening opportunities in a codebase, informed by the domain language in CONTEXT.md and the decisions in docs/adr/. Use when the user wants to improve architecture, find refactoring opportunities, consolidate tightly-coupled modules, or make a codebase more testable and AI-navigable.

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.

Glossary

Use these terms exactly in every suggestion. Consistent language is the point — don't drift into "component," "service," "API," or "boundary." Full definitions in LANGUAGE.md.

• Module — anything with an interface and an implementation (function, class, package, slice).
• Interface — everything a caller must know to use the module: types, invariants, error modes, ordering, config. Not just the type signature.
• Implementation — the code inside.
• Depth — leverage at the interface: a lot of behaviour behind a small interface. Deep = high leverage. Shallow = interface nearly as complex as the implementation.
• Seam — where an interface lives; a place behaviour can be altered without editing in place. (Use this, not "boundary.")
• Adapter — a concrete thing satisfying an interface at a seam.
• Leverage — what callers get from depth.
• Locality — what maintainers get from depth: change, bugs, knowledge concentrated in one place.

Key principles (see LANGUAGE.md for the full list):

• Deletion test: imagine deleting the module. If complexity vanishes, it was a pass-through. If complexity reappears across N callers, it was earning its keep.
• The interface is the test surface.
• One adapter = hypothetical seam. Two adapters = real seam.

This skill is informed by the project's domain model. The domain language gives names to good seams; ADRs record decisions the skill should not re-litigate.

Process

1. Explore

Read the project's domain glossary and any ADRs in the area you're touching first. If the project lacks CONTEXT.md or docs/adr/, proceed anyway — the absence doesn't block the process.

Then use delegate_task with tasks array for parallel subagents.

Rule: max 3 concurrent children. If you have 4+ exploration targets, split into sequential batches. NEVER exceed the limit — the agent will silently drop children and produce incomplete results.

# Batch 1 of 2 (3 targets)
delegate_task(tasks=[
  {goal: "Explore X", toolsets: ["terminal", "web"]},
  {goal: "Explore Y", toolsets: ["terminal", "web"]},
  {goal: "Explore Z", toolsets: ["terminal", "web"]},
])
# AFTER first batch completes, THEN start batch 2:
delegate_task(tasks=[
  {goal: "Explore W", toolsets: ["terminal", "web"]},
])

IMPORTANT: Do not jump ahead. After exploration, you MUST present candidates and wait for the user to pick. Do NOT do additional research beyond what the exploration subagents returned, do NOT propose implementation plans, and do NOT make design decisions before the user selects a candidate. The grilling loop is for the user to challenge the design with you — you are not the one making unilateral choices.

Don't follow rigid heuristics — explore organically and note where you experience 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 candidates

After subagent exploration completes, synthesize raw findings into skill vocabulary before presenting. Extract: module names, seam violations, depth problems. Do NOT dump subagent summaries verbatim — rewrite in the glossary's language (Module, Interface, Seam, Depth, Leverage, Locality).

Present a numbered list of deepening opportunities. For each candidate:

• Files — which files/modules are involved
• Problem — why the current architecture is causing friction (use: shallow, seam leak, locality failure)
• Solution — plain English description of what would change
• Benefits — explained in terms of locality and leverage, and also in how tests would improve

Use CONTEXT.md vocabulary for the domain, and LANGUAGE.md vocabulary for the architecture. If CONTEXT.md defines "Order," talk about "the Order intake module" — not "the FooBarHandler," and not "the Order service."

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 (e.g. "contradicts ADR-0007 — but worth reopening because…"). Don't list every theoretical refactor an ADR forbids.

Do NOT propose interfaces yet. Ask the user: "Which of these would you like to explore?" Stop here. Do not proceed to step 3 until the user picks. The grilling loop is for collaboratively walking the design tree — it requires a human in the loop to challenge assumptions and make tradeoffs. You do not get to skip the human.

3. Grilling loop

Prerequisite: user has picked a candidate from step 2. If they haven't, stay in step 2.

Walk the design tree with the user — constraints, dependencies, the shape of the deepened module, what sits behind the seam, what tests survive. You propose nothing unilaterally. Your job is to answer questions, surface tradeoffs, and let the user drive the design direction.

Side effects happen inline as decisions crystallize:

• Naming a deepened module after a concept not in CONTEXT.md? Add the term to CONTEXT.md — same discipline as /grill-with-docs (see CONTEXT-FORMAT.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. See ADR-FORMAT.md.
• Want to explore alternative interfaces for the deepened module? See INTERFACE-DESIGN.md.

References

Session-specific architectural findings for Rairos (and generalizable patterns) are captured in references/rairos-findings.md.