
by Weave
Weave measures engineering work done by people and AI coding agents. It combines output, quality, delivery and AI usage data to show where work gets stuck and how AI contributes.
Web appContact for pricing
Step 1: Book a demo or start on the free Starter plan.
Step 2: Connect your repositories and AI coding tools.
Step 3: Review output, quality and AI usage dashboards for your team.
Step 4: Open a representative pull request and read the output reasoning alongside the diff.
Step 5: Ask Wooly where delivery is getting stuck, in the app or through MCP.
Step 6: Optionally run npx @weave-os/router to route coding agent requests, then check status.
It is an estimate of how much work an expert engineer would need to complete a pull request. It measures effort, not recorded time or hours saved.
Yes. The Starter plan is free and covers org-level AI usage, process and quality metrics, output score, AI ROI and code turnover.
It labels evidence as verified from native tool telemetry, inferred from Git co-authorship, or estimated from activity windows. Missing evidence does not prove human-only authorship.
Yes. The quickstart describes a local Docker Compose setup and an option to point to an existing deployment, alongside the hosted setup.
GalileoGalileo is an AI observability and evaluation platform where offline evals become production guardrails. It includes 20+ out-of-the-box evals and compact Luna models for monitoring all traffic.