Skip to content

Latest commit

 

History

History
51 lines (39 loc) · 2.11 KB

File metadata and controls

51 lines (39 loc) · 2.11 KB

Testing & Quality Matrix

Testing Philosophy

TeamGraph AI uses automated project health reporting to track quality across all layers of the monorepo. Every push and pull request to master triggers a comprehensive health report.

Testing Matrix

Area What is Tested Tool CI Required Target
Unit API core flows (auth, context, MCP) pytest Yes 60%+ coverage
Integration Full signup → context → approval → MCP flow pytest + TestClient Yes Main flows pass
Type Safety Web + MCP TypeScript types tsc --noEmit Yes No type errors
Lint Python compile + ESLint compileall, eslint Yes Clean
Security Secrets, dependencies, committed files Custom scanner + npm audit Yes No critical issues
Package MCP CLI build + pack npm pack --dry-run Yes Pack succeeds
Reliability Dockerfiles, env, lockfiles Custom checks Yes All present
OSS Readiness README, LICENSE, CONTRIBUTING, SECURITY File presence Yes All present

Coverage Thresholds

Metric Current Target Notes
Statement coverage ~48% 60% Achievable by adding auth + approval tests
Build success Yes Yes API compiles, MCP builds
Security audit Monitored No critical npm high vulns tracked

Running Locally

make test           # Run API tests
make test:coverage  # Run with coverage report
make lint           # Lint all components
make build          # Build web + MCP
make security       # Security audits
make health         # Full health report

Report Outputs

After running make health, reports are generated at:

  • reports/latest/health-report.md — Markdown report
  • reports/latest/health-report.json — Machine-readable JSON
  • reports/latest/health-report.html — Styled HTML report

Raising Coverage

To improve coverage from 48% toward the 60% target:

  1. Add tests for individual routers (auth, approvals, team, brain).
  2. Add tests for the curator and context service.
  3. Add MCP endpoint-specific tests.