Testing & Validation Evidence¶
This page indexes requirements traceability, verification/validation evidence and formal user-testing evidence. For testing strategy and procedures, see Automated & end-to-end testing, User testing protocol, User testing task bank and Bug tracking and the Intermediate ingestion integrated acceptance runbook.
Requirements traceability¶
- Sprint 1 requirements traceability
- Sprint 2 requirements and rubric traceability
- Project backlog and milestone plan
Verification and validation evidence¶
Command-output summaries and sanitised verification records linked to individual issues are kept
under
evidence/validation/.
There are more than 70 dated, issue-referenced records in that folder (Markdown summaries and
screenshots); they are not duplicated here. Notable examples include:
- Issue #10 — pre-CI readiness
- Issue #44 — authorisation tests
- Issue #105 — endpoint response times
- Issue #257 — dependency-cruiser verification
- Issue #273 — accessibility and responsive-design audit
- Issue #274 — security/privacy/dependency audit
- Issue #329 — Vite/Vitest toolchain migration
- Issue #578 — repository-wide code coverage
- Issue #648 — coverage post-deployment ordering and stability
- Issue #589 — multi-season back-catalogue deployed acceptance
Browse the full validation folder for a record tied to a specific issue number.
Formal user testing¶
The canonical Sprint 2 process is the task-based protocol established for Issue #264 / PR #317 and recorded in ADR-013. Each attempted Task ID receives its own Success / Partial / Failure outcome, and findings retain task-level traceability through decision, issue/fix and retest.
Session records follow the
user-testing evidence README
and the session template. Sprint 2 sessions are retained under
evidence/user-testing/sprint-2/, with the consolidated result recorded in
sprint-2-user-testing-summary.md.
Facilitators prepare environment-specific accounts, fixtures, batches and reusable validation/reference inputs from testing/user-testing/. The pack deliberately keeps credentials out of Git and separates safe fixture-5 validation/reference data from writable success/correction scenarios that require disposable test state.
Sprint 3 continues the same protocol and evidence model through dedicated feature-level user-feedback issues #601–#607 and #612. These issues remain separate validation/evidence activities and do not automatically block linked implementation issues from closing. Implementation issues close according to their own Definition of Done and genuine technical/process dependencies. Any Cannot Begin Until list on a user-feedback issue is a testing-readiness checklist identifying functionality that must be deployed and usable before the session starts.
Sprint 3 session evidence is retained under evidence/user-testing/sprint-3/ and consolidated in sprint-3-user-testing-summary.md. Facilitator scenarios are prepared from testing/user-testing/SPRINT3_SCENARIOS.md; credentials and API keys remain outside Git.
The execution work remains split by workflow so findings can be attributed cleanly:
| Date | Session |
|---|---|
| 23 Aug 2026 | Issue #199 — public data journeys |
| 7 Sep 2026 | P03 — supplementary public/analyst evidence — retained as supplementary evidence; it is not counted as a formal task attempt because the standard facilitator metadata/outcomes were not retained. |
| 10 Sep 2026 | P01 — formal public/analyst session and P02 — formal public/analyst session. |
| 10 & 13 Sep 2026 | Issue #418 — reviewer / administrator P04 — initial REV-01 failure exposed #463; the same task succeeded after the fix, followed by REV-02, REV-05 and ADM-02. |
| 11 Sep 2026 | Issue #417 — P05 submitter — first external submitter session; accepted findings were linked to implementation work before the second session. |
| 15 Sep 2026 | Issue #417 — P06 submitter — second external submitter session; retested several P05 improvements and recorded four new findings that remain explicitly pending team disposition. |
Sprint 2 formal user testing is complete for the three tracked workstreams. The consolidated summary remains authoritative for finding decisions, issue links and retest status; unresolved or pending findings are carried visibly rather than treated as missing test evidence.
Reviewer / administration traceability¶
Issue #418 exercises the reviewer/admin workflow implemented through #283 and #362 and supplies the representative-user evidence reused by #364. The initial P04 session exposed #463 when the prepared batch remained Stored and never appeared in the review queue. #463 restored the deployed worker/Service Bus path; P04 then repeated REV-01 successfully and completed REV-02, REV-05 and ADM-02.
Related engineering defects from the same wider ingestion/review audit include #465, #471, #479, #480, #481, #482, #483, #484, #485, #486, #487 and #488. Those issues are retained as engineering context and are not attributed to P04 unless the participant independently observed them during a recorded task.
The preceding page was planned and drafted with the assistance of Claude[Claude Sonnet 5] and reviewed and edited with ChatGPT-Web[GPT-5.6 Sol], resolving issue #254 and the Sprint 2 user-testing process update.