skills/mp-architecture-review/SKILL.md
Reviews codebase architecture and opens a refactor issue with the findings.
npx skillsauth add MartinoPolo/mpx-claude-code mp-architecture-reviewInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Explore a codebase like an AI would, surface architectural friction, discover opportunities for improving testability, and propose module-deepening refactors as GitHub issue RFCs.
A deep module (John Ousterhout, "A Philosophy of Software Design") has a small interface hiding a large implementation. Deep modules are more testable, more AI-navigable, and let you test at the boundary instead of inside.
Before starting:
${CLAUDE_SKILL_DIR}/../shared/deep-modules.md now — deep vs shallow module evaluation.${CLAUDE_SKILL_DIR}/../shared/interface-design.md now — interface design rules for testability.${CLAUDE_SKILL_DIR}/REFERENCE.md now — dependency categories and the issue template.Spawn Explore sub-agent (breadth: very thorough) to navigate the codebase naturally. Do NOT follow rigid heuristics — explore organically and note where you experience friction:
The friction you encounter IS the signal.
Present a numbered list of deepening opportunities. For each candidate, show:
${CLAUDE_SKILL_DIR}/REFERENCE.md for the four categoriesDo NOT propose interfaces yet. Ask the user: "Which of these would you like to explore?"
Before spawning sub-agents, write a user-facing explanation of the problem space for the chosen candidate:
Show this to the user, then immediately proceed to Step 5. The user reads and thinks about the problem while the sub-agents work in parallel.
Spawn 3+ general-purpose sub-agents in parallel with model: "opus". Each must produce a radically different interface for the deepened module. Tell each to reason through the trade-offs before committing to a shape.
Prompt each sub-agent with a separate technical brief (file paths, coupling details, dependency category, what's being hidden). This brief is independent of the user-facing explanation in Step 4. Give each agent a different design constraint:
Each sub-agent outputs:
${CLAUDE_SKILL_DIR}/REFERENCE.md)Present designs sequentially, then compare them in prose.
After comparing, give your own recommendation: which design you think is strongest and why. If elements from different designs would combine well, propose a hybrid. Be opinionated — the user wants a strong read, not just a menu.
Create a refactor RFC as a GitHub issue. Use the template in ${CLAUDE_SKILL_DIR}/REFERENCE.md. Do NOT ask the user to review before creating — just create it and share the URL.
gh issue create --title "refactor: [module description]" --label "refactor" --body "$(cat <<'EOF'
[Use issue template from REFERENCE.md]
EOF
)"
Print the issue URL.
tools
Show current project progress. Displays phase status and next steps. Use when: "show status", "project progress", "what's done"
data-ai
Unified project setup. Auto-detects state and orchestrates mpx skills/agents for init, conversion, or restructure.
testing
Track bugs/issues in .mpx/ phase system. Parses reports, finds related phases, adds fix tasks or creates bugfix phases. Use when: "track this bug", "add issue to project", "log this bug", "add bug to checklist"
development
Project workflow guidance for spec-driven development. Background knowledge auto-loaded when relevant.