Execute a manual functional test case, updating its status dynamically and managing the test session immutability.
Installation
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Execute a manual functional test case, updating its status dynamically and managing the test session immutability.
license
MIT
metadata
{"author":"AI","version":"2.0"}
Test Execution Workflow
Follow these instructions when the user wants to execute test cases or plan a new test run session through the /test-execute workflow.
If the User wants to PLAN a new test run:
Ask them the goal, context, and which domains/tests should be included.
Create a new directory in test-management/test-runs/ following the format YYYY-MM-DD-purpose.
Create run-plan.md in this directory based on test-management/test-runs/TEMPLATE_RUN_PLAN.md.
Create a results/ directory inside the test run.
Copy the requested test cases from test-management/tests/ to the results/ directory.
For each copied test case, inject the "Execution Result" metadata block immediately after the main Title:
## Execution Result
| Who | When | Result | Where | Comment |
| :--- | :--- | :--- | :--- | :--- |
| [Name/AI] | [YYYY-MM-DD] | `⏳ PENDING` | [Context/Browser] | |
Populate the run-plan.md statistics and checklist.
If the User wants to EXECUTE a planned test run:
Locate the active run-plan.md in test-management/test-runs/.
Find the first pending test case in the checklist.
Open the copied file in its results/ directory (e.g. test-management/test-runs/2026-04-09-release/results/TC_CRT_001.md).
Present the steps to the user one by one (or in batches) through the chat interface.
Wait for the user to report passed/failed for the steps.
Check off the steps [x] inside the snapshot Markdown file.
Once the test case concludes:
Update the Execution Result block in the test case markdown.
Update the Checklist in run-plan.md to indicate ✅ PASSED, ❌ FAILED, or 🚧 BLOCKED.
Recalculate the 📊 Live Statistics table perfectly in run-plan.md.
Append an entry to the 📝 Execution Log section in run-plan.md with a summary of the result, any observations, test data corrections made, and references to any defects or improvements logged (e.g., BUS-BUG-15).
Proceed to the next pending test or yield back to the user.
During Execution — Capturing Findings
Defects & Improvements: When the user reports a bug or improvement idea during testing, log it immediately to the appropriate backlog (e.g., specifications/backlog/BUSINESS_BACKLOG.md) and note the backlog ID in the Execution Log entry for that test case.
Test Data Corrections: If test data in a template is found to be incorrect during execution (wrong locale term, wrong currency, etc.), update the template in test-management/tests/ immediately and note the correction in the Execution Log.
Scope Changes: If new test cases are created during the run (e.g., a gap is discovered), add them to the results/ directory with the Execution Result block, append them to the checklist with an ⚡ Added during execution annotation, and update the Total Scope count in the statistics table.
After Execution — Execution Report
When all test cases in the checklist are completed:
Set the run-plan status to ✅ Completed and fill in the Finished timestamp.
Create an execution-report.md in the test run directory based on test-management/test-runs/TEMPLATE_EXECUTION_REPORT.md.
Populate the report by compiling data from:
The run-plan.md statistics and execution log.
The Execution Result blocks in each results/ snapshot file.
Any backlog items created during the run.
The Execution Report is the final deliverable of a test run — a single document that captures the complete picture for stakeholders.