What problem does it solve? After a deployment, teams need a fast, repeatable way to confirm the build, endpoints, and browser surface are operational without running a full QA suite. This Skill automates that smoke-check gate and records the latest status for later comparison. ## Core Features & Use Cases - Sequential health checks: Runs build verification, E2E tests, and browser health checks declared in .autopus/project/canary.md, then issues a PASS/WARN/FAIL verdict. - Watch and compare modes: Repeats checks on a configurable interval (--watch, default 5m, max 30m) and compares results against a previous run via --compare <commit>. - Persistent status: Writes the latest verdict to .autopus/canary/latest.json so pipelines and dashboards can consume the current deployment state. - Use Case: After deploying a staging build, run /auto canary --url https://staging.example.com to confirm the build, endpoints, and browser surface are healthy before promoting to production. ## Quick Start Ask the agent to run the canary health check against your deployed URL, for example: run /auto canary --url https://staging.example.com and report the verdict.