Answers from the documentation; checking which model is on duty
Hi - I answer from the OpenSmartRoute documentation: routing, the API, plans and quotas, self-hosting. Ask away, or open a support ticket if you need a person.
Grounded in the docs - follow a source before acting on it.
gsd/verify-work - Prompt - OpenSmartRoute
Promptv1.0.0
gsd/verify-work
Guide manual user acceptance testing of recently built features
Prompt file imported from leopoldlacroix/Cognyx_case (.cursor/commands/gsd/verify-work.md). Fill in {{arguments}} before use. Copyright stays with the author.
Purpose: Validate that what Claude thinks was built actually works from the user's perspective. The USER performs all testing — Claude generates the test checklist, guides the process, and captures issues.
Output: Validation of features, any issues logged to phase-scoped ISSUES.md
Don't run automated tests (that's for CI/test suites)
Don't make assumptions about test results — USER reports outcomes
Don't skip the guidance — walk through each test
Don't dismiss minor issues — log everything user reports
Don't fix issues during testing — capture for later
</anti_patterns>
<success_criteria>
Test scope identified from SUMMARY.md
Checklist generated based on deliverables
User guided through each test
All test results captured (pass/fail/partial/skip)
Any issues logged to phase-scoped ISSUES.md (not global)
Summary presented with verdict
User knows next steps based on results
</success_criteria>
Use it
Copy one of these into your project. Installing also returns the manifest and these snippets.
Installed into a catalogue, chosen by a router
Install gsd/verify-work and it becomes one more candidate the router can pick - when it fits.
A listing is a routing target with a manifest: what it does, which domains it covers, what it costs and who publishes it. Once installed it sits beside your own models and tools, is scored like any other candidate for each request, and shows up in the trace when it wins. Ratings come from workspaces that installed it, one per account.
When a provider is down or over quota the request moves on; the trace says it did.
<objective>
Guide the user through manual acceptance testing of recently built features.
Purpose: Validate that what Claude thinks was built actually works from the user's perspective. The USER performs all testing — Claude generates the test checklist, guides the process, and captures issues.
Output: Validation of features, any issues logged to phase-scoped ISSUES.md
</objective>
<execution_context>
@./.cursor/get-shit-done/workflows/verify-work.md
@./.cursor/get-shit-done/templates/uat-issues.md
</execution_context>
<context>
Scope: {{arguments}} (optional)
- If provided: Test specific phase or plan (e.g., "4" or "04-02")
- If not provided: Test most recently completed plan
**Load project state:**
@.planning/STATE.md
**Load roadmap:**
@.planning/ROADMAP.md
</context>
<process>
1. Validate arguments (if provided, parse as phase or plan number)
2. Find relevant SUMMARY.md (specified or most recent)
3. Follow verify-work.md workflow:
- Extract testable deliverables
- Generate test checklist
- Guide through each test via AskUserQuestion
- Collect and categorize issues
- Log issues to `.planning/phases/XX-name/{phase}-{plan}-ISSUES.md`
- Present summary with verdict
4. Offer next steps based on results:
- If all passed: Continue to next phase
- If issues found: `/gsd/plan-fix {phase} {plan}` to create fix plan
</process>
<anti_patterns>
- Don't run automated tests (that's for CI/test suites)
- Don't make assumptions about test results — USER reports outcomes
- Don't skip the guidance — walk through each test
- Don't dismiss minor issues — log everything user reports
- Don't fix issues during testing — capture for later
</anti_patterns>
<success_criteria>
- [ ] Test scope identified from SUMMARY.md
- [ ] Checklist generated based on deliverables
- [ ] User guided through each test
- [ ] All test results captured (pass/fail/partial/skip)
- [ ] Any issues logged to phase-scoped ISSUES.md (not global)
- [ ] Summary presented with verdict
- [ ] User knows next steps based on results
</success_criteria>
Manifest
The prompt text and its fill-in variables. Copy it or fetch it by URL from your own code.
{
"prompt": "<objective>\r\nGuide the user through manual acceptance testing of recently built features.\r\n\r\nPurpose: Validate that what Claude thinks was built actually works from the user's perspective. The USER performs all testing — Claude generates the test checklist, guides the process, and captures issues.\r\n\r\nOutput: Validation of features, any issues logged to phase-scoped ISSUES.md\r\n</objective>\r\n\r\n<execution_context>\r\n@./.cursor/get-shit-done/workflows/verify-work.md\r\n@./.cursor/get-shit-done/templates/uat-issues.md\r\n</execution_context>\r\n\r\n<context>\r\nScope: {{arguments}} (optional)\r\n- If provided: Test specific phase or plan (e.g., \"4\" or \"04-02\")\r\n- If not provided: Test most recently completed plan\r\n\r\n**Load project state:**\r\n@.planning/STATE.md\r\n\r\n**Load roadmap:**\r\n@.planning/ROADMAP.md\r\n</context>\r\n\r\n<process>\r\n1. Validate arguments (if provided, parse as phase or plan number)\r\n2. Find relevant SUMMARY.md (specified or most recent)\r\n3. Follow verify-work.md workflow:\r\n - Extract testable deliverables\r\n - Generate test checklist\r\n - Guide through each test via AskUserQuestion\r\n - Collect and categorize issues\r\n - Log issues to `.planning/phases/XX-name/{phase}-{plan}-ISSUES.md`\r\n - Present summary with verdict\r\n4. Offer next steps based on results:\r\n - If all passed: Continue to next phase\r\n - If issues found: `/gsd/plan-fix {phase} {plan}` to create fix plan\r\n</process>\r\n\r\n<anti_patterns>\r\n- Don't run automated tests (that's for CI/test suites)\r\n- Don't make assumptions about test results — USER reports outcomes\r\n- Don't skip the guidance — walk through each test\r\n- Don't dismiss minor issues — log everything user reports\r\n- Don't fix issues during testing — capture for later\r\n</anti_patterns>\r\n\r\n<success_criteria>\r\n- [ ] Test scope identified from SUMMARY.md\r\n- [ ] Checklist generated based on deliverables\r\n- [ ] User guided through each test\r\n- [ ] All test results captured (pass/fail/partial/skip)\r\n- [ ] Any issues logged to phase-scoped ISSUES.md (not global)\r\n- [ ] Summary presented with verdict\r\n- [ ] User knows next steps based on results\r\n</success_criteria>",
"variables": [
"arguments"
],
"notes": "Guide manual user acceptance testing of recently built features Tools: Read, Bash, Glob, Grep, Edit, Write, AskUserQuestion. Arguments: [optional: phase or plan number, e.g., '4' or '04-02'].",
"metadata": {
"source": {
"provider": "github-cursor-commands",
"repository": "https://github.com/leopoldlacroix/Cognyx_case",
"path": ".cursor/commands/gsd/verify-work.md",
"ref": "aa57ce7c360a57db6f22751eb65b71a8b327ef9a",
"url": "https://github.com/leopoldlacroix/Cognyx_case/blob/aa57ce7c360a57db6f22751eb65b71a8b327ef9a/.cursor/commands/gsd/verify-work.md",
"key": "leopoldlacroix/Cognyx_case/.cursor/commands/gsd/verify-work.md"
}
}
}
Fetch it by URL: GET /api/v1/registry/leopoldlacroix-cognyx-case-verify-work-cursor-command/manifest?version=1.0.0
Reviews
Star ratings from people who tried it. One review per account; edit yours any time.
No reviews yet. Install it, try it, and be the first to rate it.
