it235 Test Design (T1 Test Design · Content + Landing)
Position
The only T1 skill: traceability → cases → full Confluence → review → (after 👤 passphrase) XMind/JIRA; parallel Apifox + automation planning.
Not for T2/T3—T2 see multica-test-t2-coverage; T3 see -t3 + multica-test-orchestration.
| Layer | Skill | Responsibility |
|---|---|---|
| Orchestration | multica-test-orchestration |
Cross-phase routing; automation into repo see references/automation-assets-lifecycle.md |
| T1 | This skill | Content + Confluence + XMind/JIRA after passphrase |
| Platform | multica-platform-* |
Read upstream; Apifox writes scenarios |
| Review | multica-review-test |
Review Confluence; review only, no import |
Three-phase delivery: Phase A is Confluence only (no XMind generated); only Phase C after 👤 passphrase does XMind + JIRA. Details in
references/generation-workflow.md.
Platform Collaboration
| Platform skill | Used for |
|---|---|
multica-platform-jira |
fetch_all → get_issue.py; after passphrase import_to_tracker.py (team-provided, see references/import-contract.md) |
multica-platform-confluence |
Collect fetch_page; publish t1-cases.md (see § Confluence landing) |
multica-platform-figma |
fetch_all → fetch_file.py |
multica-platform-apifox |
T1 parallel write interface scenarios |
| team-knowledge-base (optional) | Step 0.5 terms/entry (binary usage; not bundled in this repo, connect per platform used) |
Credentials: JIRA_USERNAME / JIRA_PASSWORD preferred → see platform skill + data-sources.md.
references/ Index
| File | Purpose |
|---|---|
generation-workflow.md |
Two phases, passphrase, test-management platform naming/validation, steps/priority, dedup/merge |
executability-and-permissions.md |
Permission identification order, dual targets, module naming, post-generation self-check |
output-format.md |
JSON 1:1, generated_summary, coverage self-check table, Output Delivery |
multimodal-content.md |
HTML/image Read and fused understanding |
agent-instructions.md |
Copyable Agent instructions |
design-trace.md |
AC ↔ design traceability |
test-case-template.md |
Functional case fields |
coverage-dimensions.md |
Dimension selection, scale, coverage matrix, XMind |
coverage-checklist.md |
Pre-generation self-check list |
web-ui-checkpoints.md |
Figma → UI test points |
api-testing-guide.md |
Interface scenario patterns |
data-sources.md |
Collection orchestration and credentials |
T1 Flow
Step 0 — Read MULTICA.md (automation path)
Issue repo matrix → each repo Read root MULTICA.md §2 (one per split repo; one per repo must be complete). Missing → BLOCKED.
Step 0.5 — Knowledge base (Step 0.5)
Team knowledge-base Q&A script (optional) → --health + Q&A. Binary usage: operational bucket goes into steps, rules bucket is not AC. See generation-workflow.md.
Step 1 — Design traceability
design-trace.md → write traceability table into coverage report.
Step 2 — Collect
python scripts/fetch_all.py --jira-url "http://jira.../browse/PROJ-123" -o ./data
Multimodal: multimodal-content.md.
Step 3–4 — Cases + self-check
By test-case-template.md + generation-workflow.md + output-format.md → local JSON (working file); self-check with coverage-checklist.md. No XMind.
Step 5 — Phase A: Confluence + Apifox parallel
- Write the full case body to
docs/test/<ISSUE-KEY>/t1-cases.md(template see platform-confluencetest-t1-template.md) - Confluence landing (see § below) → return link
- API contract ready →
multica-platform-apifoxsupplements scenarios by contract (AI branch; write manifest planning section) - →
multica-review-test(review Confluence) - Delivery note:
Status: pending human review (Confluence updated, not XMind / not imported to JIRA)
Step 6 — Phase C: XMind + JIRA (after 👤 explicit passphrase)
generate_xmind.py← latest JSON (consistent with Confluence)import_to_tracker.py(team-provided, interface seereferences/import-contract.md)- Republish Confluence status "imported to JIRA" + case-set link
Confluence Landing (T1 · master copy)
Parent page: JIRA Issue Hub (PRD requirement page) child page.
Local draft: docs/test/<ISSUE-KEY>/t1-cases.md
Baseline: multica-platform-confluence/scripts/templates/test-t1-template.md
Must include each CASE step/expected body, Apifox section, UI e2e plan (path in MULTICA.md §2, not hardcoded). T1/T2 do not generate XMind.
Publish:
python multica-platform-confluence/scripts/publish_design.py \
<ISSUE-KEY> docs/test/<ISSUE-KEY>/t1-cases.md \
--append-jira --json
JIRA block: h3. T1 Test Cases (Test Cases T1).
Apifox (T1 parallel)
After contract published: read API contract → multica-platform-apifox supplements scenarios (AI branch) → Confluence records branch/scenario → manifest path in MULTICA.md §2.
Files
multica-test-t1-design/
├── SKILL.md
├── config.local.env.example
├── references/
└── scripts/
Why it works
T1 content and scripts are unified; generation-workflow + output-format preserve the original generator's hard rules; platform uniformly collects via REST.