TEA Audit

What It Is

The TEA audit is the testing-stage quality gate in the software factory pattern. It evaluates completed implementation against requirements across three dimensions—Test (functional correctness), Evaluate (non-functional properties), and Assess (overall quality)—and produces a structured decision about whether work is ready for release.

Key Patterns / Components

Three Evaluation Dimensions

T — Test (Functional)
- Does the implementation meet the stated requirements?
- Are all features working as specified in the PRD?
- Edge cases and error conditions handled?

E — Evaluate (Non-Functional)
- Performance under load
- Security posture (no hardcoded credentials, SQL injection vectors, etc.)
- Scalability considerations
- Operational readiness

A — Assess (Quality)
- Code quality (function length, error handling, documentation)
- Test coverage
- Documentation completeness
- adherence to architecture decisions

Gate Outcomes

  • PASS — Ready for release, no blocking issues
  • PASS-WITH-FOLLOWUPS — Releaseable but with known issues to address post-launch
  • REMEDIATE — Blocking issues must be fixed before release

TEA Summary Output

The test-lead agent outputs tea-summary.md documenting:
- Test results per requirement
- Non-functional evaluation findings
- Quality assessment scores
- Gate decision with rationale
- Action items for follow-ups (if applicable)

How It Applies to Marketing Factory

A content QA audit follows an analogous structure: Test (does the content match the brief?), Evaluate (does it meet quality standards—SEO, readability, brand voice?), Assess (is it publication-ready?). A content audit might gate content as APPROVED, APPROVED-WITH-REVISIONS, or RETURN-TO-DRAFT.