Repo Watch

Huge thank you to everyone sharing feedback with me over email and other channels. It’s directly shaping Repo Watch. Since the last update, we’ve shipped a lot: - Compare scans with baseline score tracking so you can see drift scan-to-scan - Support for user-provided test coverage artifacts across different language stacks - “Intentional” finding tracking for false positives in your own context across scans - Next iteration of downloadable reports (PDF + Markdown), now including score drift between scans - Free tier update: all individual findings are now visible, not just grouped findings There’s much more coming. The long-term vision for Repo Watch is simple: “I inherited this repo. Tell me what to fix first this week, and prove whether risk is moving up or down over time.” https://repowatch.io/releases

In my latest update i share how we got up in running over just 9 days. https://repowatch.io/blog/repowatch-week-one-what-we-built You can read about all the features we released and how the latest Complexity Risk Score works. How Complexity Risk Score works Complexity Risk measures how likely a codebase is to have performance scaling problems — without executing any code. The scanner reads every source file and uses pattern matching to detect common code shapes that tend to slow down as data grows: Nested loops (loop inside a loop) — flagged as O(n²), "grows fast" Linear search inside a loop (.find/.filter/.includes in a loop body) — also O(n²) Sorting inside a loop — O(n log n), "moderate growth" Network/DB calls inside a loop (fetch, prisma, etc.) — O(n), scales linearly with data Each detected pattern gets a fixed risk rating (72–92), a complexity class label, a 4-line code snippet, and a language-specific refactoring suggestion. The top 3 worst findings drive the score — starting from a perfect 100 and deducting 6–20 points each based on severity, with a small extra penalty if more than 3 hotspots exist. When nothing is found, the panel flips to green "all clear" signals for each pattern category. Complexity Risk contributes 10% of the overall repo health score, alongside Code Quality (30%), Test Confidence (25%), Security Hygiene (25%), and AI Risk Indicators (10%).