How We Build / Automated Test Report
Cloud Spectra sits in the data path of critical traffic, so correctness is engineered, not hoped for. The numbers below are measured directly from the codebase -- not rounded up for effect.
12,947+
automated unit & component tests
10,598+
Go backend test functions
2,349+
Angular specs across 164 suites
89.7%
Go backend statement coverage
What Runs on Every Build
A single command runs all of these suites in sequence; a change cannot become a release unless every one is green. Individual test names and code are deliberately omitted -- this is a sanitized, public view.
| Suite | Scale | Result |
|---|---|---|
| Backend unit tests (Go) Packet rules, policy generation, config merges, teardown logic -- run under coverage. | 10,598+ functions | PASS |
| Dashboard specs (Angular) Component logic, service methods, and every subscribe branch and state transition. | 2,349+ specs | PASS |
| Emulated integration (moto) Resource ordering and control-plane dependencies against a faithful AWS-API emulator. | -- | PASS |
| Deploy-template lint (cfn-lint) CloudFormation templates validated before they can ship. | -- | PASS |
| Terraform validate + provider mirror Native Terraform module and provider mirror validated for the parity lane. | -- | PASS |
| Data-plane fuzz tests (ASan) Adversarial packet-parser fixtures for the fast-path, built under AddressSanitizer. | -- | PASS |
| Build & release tooling Schema codegen, pricing/catalog consistency, release channels, and packaging guards. | -- | PASS |
Backend Code Coverage
Coverage is measured on every build over hand-written Go (trivial main() wrappers and generated API clients are excluded, so the number reflects real logic).
89.7%
of 46,061 statements covered · 39 packages · 383 source files
24
packages >= 90%
9
packages 80-90%
1
packages 70-80%
5
packages < 70%
Generated by the build, not by hand. This report is produced by the test pipeline itself and is published only when the entire suite passes. If any suite fails, the build stops before this page is ever regenerated.
Sanitized on purpose. No source code, file paths, or internal identifiers appear on this page. Its content is limited to aggregate counts, coverage percentages, and human-written category labels, and an automated guard fails the build if anything else slips in.
See Also
This page covers the unit and emulated tiers. The top of the testing pyramid -- full stacks on real AWS -- is summarized separately.
Automatically generated from the codebase on 2026-07-16. Figures are floor counts measured from the codebase; live subtests and parameterized cases push the real totals higher.