Today, I’m launching Hotspots — a lightweight way to surface the tiny set of functions that dominate your risk and to keep them from getting worse.
What’s shipping
- Snapshot analysis: rank functions into risk bands and generate an optional HTML report for sharing.
- Explain mode: driver labels that clarify why a function is risky and where to focus a refactor.
- CI guardrails: delta checks and policies that block risky regressions before they merge.
- Simple CLI: no build step required; point it at your repo and go.
Why I built it
Refactors often start with “this looks messy,” then balloon without moving the needle. Incidents, on the other hand, keep recurring in the same few places — complex code that also changes often. I built Hotspots to make those places obvious, actionable, and protected by policy.
How it fits your day
- Local triage: run a snapshot, scan the ranked list, and pick the top 2–3 items.
- PR safety net: run delta mode in CI so complexity spikes fail early.
- Shareable context: attach an HTML report to planning docs or retros.
Quick start
hotspots analyze src/
hotspots analyze src/ --mode snapshot --format html --output .hotspots/report.html
hotspots analyze src/ --mode delta --policyWhat you’ll see
- A prioritized list of functions with risk bands and clear driver labels
- Concise actions (e.g., “Reduce branching; extract sub‑functions”) to guide refactors
- Policy results for PRs with pass/warn/fail signals
Roadmap (next few iterations)
- Richer policy presets and easier tuning per repo
- Better suppression/waiver workflow with reasons and expiry
- Trend views to track counts of critical/high functions over time
Docs and examples live at https://docs.hotspots.dev. Source at https://github.com/Stephen-Collins-tech/hotspots.
Refactor what matters — and keep it that way.