Hotspots

Multi-language complexity analysis — single binary, zero config, results in seconds.

Hotspots finds the small share of functions responsible for most of a codebase's review burden: the ones combining high structural complexity with recent, frequent change. It scores every function, ranks them, and — in a git repo — layers in churn and touch frequency to separate active risk from dormant debt.

Get started → · CLI reference → · GitHub →

What it does

• LRS scoring — combines cyclomatic complexity, nesting depth, fan-out, and non-structured exits into one actionable number per function.
• Git-enriched triage — snapshot mode adds churn and touch frequency, placing every function in a fire / debt / watch / ok quadrant. Call-graph metrics (fan-in, PageRank, betweenness) are computed and reported alongside, but don't feed the live score.
• Policy engine — block PRs that introduce critical-risk functions or regress LRS. Works with GitHub Actions, GitLab CI, and any CI that checks exit codes.

Where to go next

Quick Start	Install the binary and run your first analysis.
Usage & Workflows	Snapshot mode, delta mode, the policy engine, hotspots coordinate.
CLI & Config Reference	Every command, every flag, every config field.
Architecture	How the analysis pipeline actually works, end to end.
Score Stability & Predictability	What "the score didn't change for no reason" actually means and why it holds.
Contributing	Dev setup, adding a language, release process.

Live analyses of real open-source repositories run continuously on hotspots.dev — see what this looks like against a codebase you already know.