02 flagship station · ai agent
TestScope AI
Reads a GitHub issue, reads your pytest suite, tells you what isn't tested.
Give it a repo and an issue number. A LangGraph agent pulls the acceptance criteria, analyzes the existing tests through a custom MCP server and the GitHub MCP server, and returns a coverage matrix with missing-test recommendations. If you approve, it files a GitHub issue for the gaps.
- Jobs are queued on SQS and consumed by a separate worker
- Self-managed Kubernetes on AWS, provisioned with Terraform
- GitHub Actions CI/CD · Prometheus / Grafana / Loki