improve-codebase-architecture

Identify evidence-backed architectural deepening opportunities in codebases and route selected improvements through PREVC.

Updated Jul 29, 2026
One-click install
npx skills add https://github.com/MaiconGambini/opencode-harness-guide --skill improve-codebase-architecture-maicongambini
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: improve-codebase-architecture
Source: https://github.com/MaiconGambini/opencode-harness-guide/tree/main/skills/improve-codebase-architecture
Command: npx skills add https://github.com/MaiconGambini/opencode-harness-guide --skill improve-codebase-architecture-maicongambini

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Codebases accumulate shallow modules, duplicated policy, and hidden coupling that make changes risky and testing difficult. This Skill finds targeted, evidence-backed architectural improvements in code that changes frequently or resists testing, instead of producing broad speculative refactor lists. ## Core Features & Use Cases - Evidence-Based Review: Scopes analysis to a user-named area or recent hotspots from git history, grounded in CONTEXT.md and existing ADRs. - Structured Candidate Reports: Presents each opportunity with affected areas, observed problem, smallest deepening direction, testing benefit, risk, and recommendation strength. - PREVC Routing: After the user selects a candidate, routes the change through spec-lead requirements clarification and the PREVC planning-execution-validation workflow. - Use Case: A developer notices a payment module is painful to modify. The Skill reviews recent churn, identifies a shallow pass-through layer and duplicated validation policy, and proposes deepening the module behind a testable seam before any code is written. ## Quick Start Ask the agent to review the checkout module for architectural deepening opportunities using recent git history as evidence.

Frequently Asked Questions about improve-codebase-architecture

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I find architectural improvement opportunities in a codebase?

Scope the review to a named area or recent hotspots from git log, then look for shallow pass-throughs, scattered knowledge, duplicated policy, and hidden coupling. Each finding must be backed by concrete code evidence, not speculation.

What is the deletion test in refactoring?

The deletion test asks whether a module or abstraction could be removed without losing real behavior. It filters out refactor proposals that only reshuffle code, ensuring candidates represent genuine deepening of module boundaries.

Does this Skill implement the refactor automatically?

No. It only identifies and presents candidates with risks and recommendations. Implementation begins only after the user selects a candidate, which is then clarified with spec-lead and routed through the PREVC workflow.

When should I not use architecture review skills?

Avoid using it to generate broad speculative refactor lists or to overturn existing ADRs without concrete evidence. It is designed for targeted, evidence-backed deepening in high-churn or hard-to-test areas, not wholesale redesign.

What inputs does an architecture review need?

The review reads CONTEXT.md and relevant ADRs first, plus git log output to locate recent hotspots. A user-named area can also scope the analysis to a specific module or subsystem.