improve-codebase-architecture · Skill Details

Mechanism

When It Triggers · How It Works · What It Produces

Breaks down how the skill works and what it produces, so you can quickly judge whether it fits your scenario and value.

Trigger: User-invoked during spare moments for codebase health improvement. /ask-matt classifies it under "Codebase health." Also recommended by /diagnosing-bugs when structural issues are found.
Mechanism: Three steps: (1) Explore — identify hot spots via git log, use an Explore sub-agent to scan the codebase, apply the deletion test and other methods to find shallow modules; (2) Generate an HTML report — for each deepening candidate, provide "Problem → Solution → Benefits → Before/After diagrams → Recommendation strength", write to OS temp directory and open in browser; (3) Grilling — once the user picks a candidate, start /grilling to walk the decision tree, invoke /domain-modeling to update the domain model, and use /codebase-design for the "design-it-twice" pattern when needed.
Output: An on-the-fly HTML architecture review report and an actionable refactoring plan (confirmed via grilling). Candidates conflicting with existing ADRs are flagged with justification for reopening.

Use Cases

Two Typical Use Cases

Scenario 01

Periodic codebase architecture review

A tech lead runs `/improve-codebase-architecture` every few days to check codebase health. The skill identifies hot spots via git log, uses /codebase-design vocabulary to find shallow modules, and generates an HTML report with before/after diagrams. The lead picks a candidate, and the skill starts a /grilling loop to walk the design decision tree, updating the domain model via /domain-modeling.

Scenario 02

Architecture improvement discovered after bug fixing

After fixing a bug where no suitable seam existed for a regression test (recorded by /diagnosing-bugs), the developer switches to `/improve-codebase-architecture`. The skill scans the relevant modules to identify root architectural issues causing the missing seam, generates a report with refactoring options, and guides through specific design decisions once the user picks one.

Skill Relationships

Which Skills I Use · Which Skills Use Me

Dependency relationships read as "the upper tier points to the lower tier." The current Skill sits in the middle tier — above are Skills that depend on it, below are Skills it depends on.

Tier 1 · These Skills Use Me

Tier 2 · Current Skill

improve-codebase-architectureWorkflow

Tier 3 · I Use These Skills

Full Relationship Graph

Browse this skill's relationships within its skillset. Click a node to switch the side panel; use the search box to jump to any skill.

Skill File

SKILL.md

Open in GitHub ↗

SKILL.mdLoading from GitHub…

Skill Composition

Open in GitHub ↗

SKILL.md33.3%
other66.7%

▸skills/improve-codebase-architecture/
▸improve-codebase-architecture
- ·HTML-REPORT.md
- ·SKILL.md
- ▸agents
  - ·openai.yaml

Per the Agent Skills spec, SKILL.md is the only required file. The colored bars classify the rest into the three spec-aligned optional directories (scripts / references / assets) and an 'other' bucket. The full directory tree is below.

Related in Skillset

You May Also Need These Skills

ask-matt

A router skill for the entire skillset: when the user can't remember which skill to use or which flow fits their situation, they invoke this to get a recommendation. It describes the main "idea → ship" flow (grill-with-docs → to-spec/to-tickets → implement), plus on-ramps (triage, diagnosing-bugs, wayfinder), codebase health (improve-codebase-architecture), vocabulary layers (domain-modeling, codebase-design), and standalone skills (grill-me, prototype, research, teach). Following its map leads users to the correct workflow entry point.

Workflow

code-review

A two-axis code review skill: reviews a diff (from a fixed point to HEAD) along both Standards (coding standards + Fowler smell baseline) and Spec (faithful implementation of the originating issue/PRD). Both axes run as parallel sub-agents to avoid context pollution. The report aggregates findings without merging conclusions, ensuring "passes standards but implements the wrong thing" and "implements the right thing but breaks conventions" are both surfaced independently.

Workflow

codebase-design

A shared vocabulary skill for deep module design: provides a precise set of design terms (Module, Interface, Depth, Seam, Adapter, Leverage, Locality) guiding agents and developers to design "deep modules" — a lot of behaviour behind a small interface. Includes design principles (the deletion test, the interface is the test surface) and testability guidelines. A reference skill consumed by other skills (e.g., improve-codebase-architecture, tdd) and invoked directly by engineers.

Style Guide

diagnosing-bugs

A disciplined diagnosis loop for hard bugs and performance regressions. The core idea is to first build a tight feedback loop — no hypothesis phase without one. Six phases: build feedback loop → reproduce + minimise → generate 3-5 falsifiable hypotheses → instrument by prediction → fix + regression test → cleanup + post-mortem. If the codebase architecture can't lock down the bug, it recommends handing off to the improve-codebase-architecture skill.

Workflow