Testing Ticket (discovery stub)

Thin entry for agent skill discovery. Full procedure: WORKFLOW.md.

When invoked:

1. Announce once: Using **testing-ticket-workflow** to test the ticket with Playwright.
2. Settle and strategize first — before any other tool call (the Dependencies check below runs right after this step, before Phase A), follow settle-and-strategize.md: invoke `engineering:testing-strategy` to plan the approach, no guessing an expected/spec, no blind retry loops. If that skill is not in the session's available-skills list, write the strategy inline (scope, order, done-when, risks) and say in chat that you did so (dependencies.md).
3. Design (Figma) comparison on every UI scenario — figma-design-comparison.md; rule stated once at WORKFLOW Phase B step 4 and E0.
4. Test deeply — follow customer-escape-prevention.md; runs at WORKFLOW Phase B step 5, E0 and QA closing Depth gates.
5. Parallel lanes by default, one leased account per lane — parallel-test-lanes.md; lane plan at WORKFLOW Phase B step 7, dispatch and lane failure in Phase E.
6. Login by command, never a login window or a password — when an account must log in or its session is dead, give the user ready-to-run commands as one combined bash block per step that loops over the accounts (each isolated, prints OK / FAILED with a log path): relink every needed account, then setup/mint only the FAILED ones; the user runs them, then probe the session and report. Never ask for, collect or fill a password. Full steps: WORKFLOW Phase D Fresh login (parallel-test-lanes.md §3).
7. Result quality gate, all 20 layers on every case before any verdict — result-quality-gate.md; mechanical layers by checker, judgement layers by hand, gate time counted in the round line (execute-and-summarize Result quality gate).
8. No-wait decision gate, all 20 layers before any BLOCKED / "needs decision" / held post — no-wait-decision-gate.md; runs at E3 (execute-and-summarize); never park a case or a finished post waiting for someone.
9. 🔴 GATE: confirm before the first Playwright run — stated once at WORKFLOW Phase C (bot-run branch included).
10. Read and follow WORKFLOW.md end-to-end — every step, gate, and reference.

Claude Code shortcut: /testing-ticket → commands/testing-ticket.md.

Contents

The must-read files for this workflow, linked directly from here (references already linked above are not repeated; WORKFLOW.md links any further file at the step that uses it). Read the Always tier at the start; read each On trigger file only when its trigger is reached.

Always (at the start of every run)

• WORKFLOW.md — always: the full procedure, read end to end.
• shared-preamble.md — at the start of every run, before the first reply (preamble, constraints, tone).
• helix-session-constraints.md — read once at session start.
• user-communication.md — read before any message to the user.
• workspace-guide-discovery.md — on the first response, to locate the workspace project guide for this workflow.

On trigger

Grouped by the phase that reaches the trigger; each line names its exact trigger.

#### Intake and plan (Phases A–C)

• session-intake.md — at Phase A — the six intake fields and the confirmation gate.
• intake-one-pager.md — at intake, to show the one-page list of inputs the run needs.
• parallel-prep.md — at Phase B, when the Jira fetch and Confluence/Swagger reads are independent.
• test-execution-plan-template.md — when writing the test execution plan shown before the run.
• skill-rules-style.md — read before asking the user or stopping on a missing input.

#### Run (Phases D–E)

• round-time-contract.md — before logging in or reusing a session for the round.
• playwright-preflight.md — before starting the Playwright run — must end with Ready to run: YES.
• playwright-discipline.md — before writing or running any Playwright step.
• resilient-selectors.md — read when writing or fixing Playwright selectors.
• test-through-real-steps.md — before executing any case — drive the real screen, never a backend shortcut.
• execute-and-summarize.md — before Phase E — read and follow Phase E, the result quality gate and Phase F, end to end.
• qa-evidence-gates.md — read when recording or judging evidence clips and the AC/EC coverage gate.
• qa-debug-discipline.md — when results conflict with expectations, with dev claims, or look flaky.

#### Verdicts and defects (Phases E–F)

• root-cause-investigation.md — before calling any case FAILED or BLOCKED — investigate to a verified root cause.
• non-pass-challenge-gate.md — before recording any non-PASSED verdict.
• bug-priority-matrix.md — whenever a Priority/Severity is named or judged.
• defect-report-completeness.md — for any verdict that is not a clean PASSED — what the write-up must contain.
• defect-candidates.md — before reporting any defect found in the round — follow this exact table and the fixed owner question; never file without the owner.

#### Writes outside the chat (Phase G)

• results-update.md — before Phase G — read and follow it whenever results are updated outside this chat.
• result-update-discipline.md — read before updating a results destination.
• shared-must-never.md — before any write or post — rules shared by every workflow.
• jira-wiki-vs-markdown.md — before and after posting any Jira body (v2 wiki vs v3 ADF).
• jira-linebreak-conversion.md — read before posting a Jira table with multi-line cells.
• jira-comment-post-review.md — read after posting a Jira comment, before saying done.

#### Close-out

• qa-closing-shared.md — before reporting done — shared QA closing gates, including the cross-ticket conflict check.
• proactive-qa-triggers.md — read at close-out, before naming the next workflow (e.g. Next: create-bug for {ids} when F3 lists a defect).
• session-closing.md — read at session end.

#### Any phase

• agent-rationalizations.md — read when tempted to skip a step; each excuse has its correct action.
• long-workflow-todos.md — read when the workflow has many steps; keep the todo list.
• skill-invocation-discipline.md — read before invoking or switching skills.
• skill-routing.md — read when choosing which skill handles a request.

Refusal-first (precondition gate)

All preconditions and refusal rules are in WORKFLOW.md. MUST NOT run Playwright until intake is complete and the WORKFLOW Phase C confirm gate has passed.

QA closing (mandatory before "done")

All close-out gates are in WORKFLOW.md and verify-closing-checklist.md (Testing ticket section). MUST NOT claim tests passed without fresh runner output, without a design node (or its BLOCKED-with-reason) on every UI row, or with passed rows that carry no evidence.

Dependencies

Run the check for each of these after step 2 (strategy) and before Phase A — one order: strategize, then this check, then intake; check commands and install hints are in dependencies.md:

• Playwright (in the project under test)
• Node.js + in-repo guard tools (tools/lane-lease-check, tools/adoption-check)
• Figma
• browser automation MCP
• QA capture and verify tools ({QA_TOOLS_ROOT})
• Python 3
• ffmpeg / ffprobe
• curl + Jira REST credentials
• engineering:testing-strategy
• superpowers:systematic-debugging