execute-and-report-journey

name	execute-and-report-journey
description	Executes a Journey test, validates UI, captures screenshots, and maintains a live Markdown report of its progress with Before/After image comparisons.
version	1.0.1
author	dyor
import_commit	af60f004d289c9611894328e0099b4956e7b1efe
import_date	"2026-05-02T08:18:11.000Z"
import_url	https://github.com/dyor/skills/tree/main/execute-and-report-journey

execute-and-report-journey Skill

When a user invokes you with "run execute-and-report-journey on <journey_name>" or a similar request, this skill executes the journey test and generates a persistent Markdown change review document with Before/After screenshots. It minimizes IDE permission interruptions by batching file writes and image processing.

Workflow

1. Initialize Directories & Report:

   • Ensure the base directory structure exists for the current journey.
   • Clear out any previous run data in the latest folder: rm -rf .journey_reports/<journey_name>/latest/
   • Create the necessary directories: mkdir -p .journey_reports/<journey_name>/latest/images/

2. Execute Step & Capture Screenshots: For each action in the journey:

   • Determine the current UI state using ui_state.
   • Perform necessary UI actions using adb_shell_input or other tools.
   • Verify the outcome using ui_state.
   • Capture Screenshot: Take a screenshot of the resulting screen by running adb shell screencap -p > .journey_reports/<journey_name>/latest/images/step_<step_number>.png.

3. Judge Action & Failure Loop:

   • Keep track of the pass/fail status and detailed grading for each step in your memory to build the report.
   • If a step fails:
     • Do NOT change the journey test to force it to pass.
     • You must attempt to repair the application code to fix the bug with at least one retry.
     • After applying the code fix, redeploy the app and restart the journey.
     • If the journey passes after a fix: When generating the final report, the screenshots from the failed/previous run will act as the "Before" images, and the screenshots from the newly successful run will act as the "After" images.
     • If the journey still fails after multiple repair attempts (e.g., 2-3 tries), you must stop work completely and ask the user for manual intervention. Do not proceed to generate a "passing" report.

4. Repeat: Continue to the next action and repeat steps 2 and 3 until all actions are completed. To avoid IDE prompt interruptions, do NOT write to the report file piecemeal or use shell commands for formatting.

5. Batch Resize Images & Finalize Report:

   • Resize all captured screenshots at once using the provided bash script to prevent constant IDE permission prompts: bash .skills/imported-skills/execute-and-report-journey/scripts/copy_and_resize_images.sh .journey_reports/<journey_name>/latest/images .journey_reports/<journey_name>/latest/images 300
   • Find Previous Archive: Use ls -td .journey_reports/<journey_name>/archive/* 2>/dev/null | head -1 to find the most recent previous run (if any).
   • Construct Markdown Report: Build the entire final markdown report in memory. For each step's screenshot, if a previous run exists, embed the images in a Before/After table:

     | Before | After |
     |---|---|
     | ![Before](file://<absolute_path_to_archived_step_image>) | ![After](file://<absolute_path_to_latest_step_image>) |
     

     If no previous run exists, just use a standard image tag: ![Step <step_number>](file://<absolute_path_to_latest_step_image>)
   • Use the native write_file tool to save the complete report.md file once. Do NOT use echo via run_shell_command as this breaks formatting and prompts the user repeatedly.
   • Generate a timestamp (e.g., yyyy-MM-dd-HH-mm).
   • Copy the entire latest/ directory to the archive: mkdir -p .journey_reports/<journey_name>/archive/<timestamp> && cp -R .journey_reports/<journey_name>/latest/* .journey_reports/<journey_name>/archive/<timestamp>/

Example Report Format

# Journey Report: 04_lines_directory.journey.xml

## Execution Summary

| Action | Status |
| :--- | :--- |
| Verify the list of transit lines is visible. | ✅ Passed |

## Detailed Grading

### 1. Verify the list of transit lines is visible.
*   **Result:** ✅ Passed
*   **Grading:** Verified list is visible and styled properly.

| Before | After |
|---|---|
| ![Before](file:///Users/mattdyor/metropolist8/.journey_reports/04_lines_directory/archive/2026-04-30-10-37/images/step_1.png) | ![After](file:///Users/mattdyor/metropolist8/.journey_reports/04_lines_directory/latest/images/step_1.png) |