From dominion-flow
Guides the user through conversational UAT testing by loading phase context, generating test flows from must-haves, and running parallel debug agents on failures.
How this command is triggered — by the user, by Claude, or both
Slash command
/dominion-flow:fire-verify-uatFiles this command reads when invoked
The summary Claude sees in its command listing — used to decide when to auto-load this command
# /fire-verify-uat > Conversational UAT testing with automatic parallel diagnosis when flows fail. --- ## Arguments --- ## Process ### Step 1: Load Phase Context Read all RECORD.md files from the phase. Extract: - What was built (from quick summaries) - Critical flows (from PROJECT.md) - Must-haves (from BLUEPRINT.md frontmatter) ### Step 2: Generate Test Flows Create test flows from must-haves truths: Present to user: "Ready to start UAT? I'll guide you through each flow." ### Step 3: Guided Testing For each flow: 1. Present exact steps to test 2. Wait for user result (pa...
Conversational UAT testing with automatic parallel diagnosis when flows fail.
arguments:
phase:
required: false
type: string
description: "Phase to test (e.g., 03). Defaults to current phase."
+---------------------------------------------------------------+
| DOMINION FLOW >>> UAT VERIFICATION |
+---------------------------------------------------------------+
Read all RECORD.md files from the phase. Extract:
Create test flows from must-haves truths:
Critical Flows to Test:
1. [Must-have truth 1] - Must Pass
2. [Must-have truth 2] - Must Pass
3. [Edge case from PROJECT.md] - Should Pass
Present to user: "Ready to start UAT? I'll guide you through each flow."
For each flow:
+---------------------------------------------------------------+
| TESTING: Flow 1 of N |
+---------------------------------------------------------------+
| |
| Flow: User can register with email and password |
| |
| Steps: |
| 1. Run: npm run dev |
| 2. Visit: http://localhost:3000/register |
| 3. Enter email: [email protected] |
| 4. Enter password: SecurePass123! |
| 5. Click "Register" |
| |
| Expected: Redirect to /dashboard, welcome message |
| |
|-----------------------------------------------------------------|
| Result? Type "pass", "fail", or describe the issue |
+-----------------------------------------------------------------+
When a flow fails, immediately spawn 3 parallel debug agents:
// Agent 1: Component where symptom appears
Task({ subagent_type: "fire-debugger", prompt: "Investigate [failing component]..." });
// Agent 2: What triggers it (parent/caller)
Task({ subagent_type: "fire-debugger", prompt: "Investigate [trigger component]..." });
// Agent 3: Working reference pattern
Task({ subagent_type: "Explore", prompt: "Find working pattern for [similar feature]..." });
Present diagnosis results and proposed fix.
For each failure with diagnosis:
fix({phase}): resolve UAT failure - [description]## UAT Report: Phase XX
### Results
| Flow | Status | Notes |
|------|--------|-------|
| Registration | PASS | |
| Login | PASS | |
| Dashboard | FAIL -> PASS | Fixed: missing redirect |
### Verdict: [PASS / CONDITIONAL PASS / FAIL]
- Blocking issues: [N]
- Fixed during UAT: [N]
- Deferred: [N]
| Verdict | Next Action |
|---|---|
| PASS | /fire-transition to complete phase |
| CONDITIONAL PASS | User decides: proceed or fix |
| FAIL | Fix blockers, rerun /fire-verify-uat |
@templates/UAT.md@references/verification-patterns.mdnpx claudepluginhub thierryn/fire-flow/verify-workValidates built features through conversational UAT, producing a phase-UAT.md report. Automatically diagnoses issues, plans fixes, and prepares for execution.
/flow-test-strategy-executionOrchestrates test strategy execution for unit/integration/e2e/regression tests, validating coverage, triaging defects, and analyzing regressions.
/test-runExecutes test cases via playwright-cli interactive browser agent, tracks bugs in bugs/BUG-NNN.md, auto-fixes P0/P1 bugs and retests until exit criteria met.
/qaCreates test plans and executes comprehensive QA testing including auth flows, edge cases, regressions, and bug reports.
/verifyVerifies implementations or changes using parallel specialized agents with 0-10 grading, full reports across tests/security/quality, and improvement suggestions.