CodeScene's structural maintainability analyzer as an MCP server. It scores files 1-10 before you touch them and verifies the delta after each change, so you know when AI-generated refactors are making things worse. The workflow is straightforward: review for a baseline, edit, score again, gate the commit if health regressed. It catches design-level problems (complexity, brain-heavy functions) that linters miss. You'll want this when refactoring legacy code or before opening a PR on a module you don't trust yet. Standalone mode works without a paid CodeScene account. The four gating tools (review, score, pre-commit safeguard, change set analysis) are what matter. Complements lint and test loops with a structural regression check.
npx -y skills add affaan-m/ecc --skill codehealth-mcp --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
sickn33/antigravity-awesome-skills
moizibnyousaf/ai-agent-skills
github/awesome-copilot