// the find
aws-solutions-library-samples/cloud-intelligence-dashboards-framework
Command Line Interface tool for Cloud Intelligence Dashboards deployment
The CUDOS Framework is AWS's official tooling for deploying a fleet of cost and usage QuickSight dashboards on top of CUR2 data via Athena. It's for FinOps teams or platform engineers at organizations with meaningful AWS spend who want something beyond the native Cost Explorer. The CLI and CloudFormation path both work; the real product is the 20+ pre-built dashboard definitions that ship with it.
The dashboard catalog is genuinely broad — CUDOS, KPI, Graviton, Compute Optimizer, Trusted Advisor, and more are all here as YAML definitions, so you're not hand-crafting QuickSight assets from scratch. The plugin architecture (cid/plugin.py, dashboards/catalog.yaml) means third parties can add dashboards without forking. CI includes bats integration tests that actually deploy and delete real dashboards in AWS, not just unit tests against mocks. Active maintenance is evident: the repo pushed yesterday and individual changelogs per dashboard make it possible to track what changed without reading a monolithic diff.
QuickSight Enterprise is a hard dependency and it runs $72–$120/month before you factor in Athena scan costs — this is not a cheap observability option for small teams. The Terraform support is marked 'legacy' and lives in a separate folder with a README warning; if you're IaC-first and not using CloudFormation, you're a second-class citizen. Test coverage in Python is thin: three test files covering CSV parsing and parameter isolation, nothing testing the actual QuickSight API interaction paths. The separation between 'foundational' and 'advanced' dashboards requires deploying a separate Data Collection stack via Step Functions, which adds significant setup complexity that the README undersells.