Skip to content

fix(vscuse): Auto-fix DA_Oauth_ts_Local_Debug (progress 39→55 steps) - #16498

Open
Tian Yuan (xiaolang124) wants to merge 8 commits into
devfrom
auto-fix/DA_Oauth_ts_Local_Debug/2026-08-01
Open

fix(vscuse): Auto-fix DA_Oauth_ts_Local_Debug (progress 39→55 steps)#16498
Tian Yuan (xiaolang124) wants to merge 8 commits into
devfrom
auto-fix/DA_Oauth_ts_Local_Debug/2026-08-01

Conversation

@xiaolang124

Copy link
Copy Markdown
Contributor

VscUse Auto-Fix Results

Plan: DA_Oauth_ts_Local_Debug
Status: ❌ FAILED (8/8 attempts exhausted)
Initial Report: Link
Final Report: Link

Iteration Summary

# Fix Applied Result Report
1 Removed the stale dhash preconditions on step_6f17d0b9 (click "M365 Agents Toolk ❌ Failed Report
2 Aligned the MS sign-in "Next" button click (step_f1471ec9 in group__debug_in_cop ❌ Failed Report
3 Cleared the stale dhash preconditions on step_db5979ea ("Declarative Agent" clic ❌ Failed Report
4 Added delay: 3 to the second-login email TYPE step (step_0fec5113) in group__d ❌ Failed Report
5 Broadened final assertion step_670cbf14 in group__debug_in_copilot_local.json fr ❌ Failed Report
6 Removed the stale dhash preconditions on step_b4728564 ("Click the input box in ❌ Failed Report
7 Removed the stale dhash precondition (dhash:512:384:0:20:11dc6346f8e6e4e8) on ❌ Failed Report
8 Aligned second-login email-field click (step_0115ed3f in group__debug_in_copilot ❌ Failed Report

⚠️ AI could not fix this plan automatically. Human review needed.
Each commit represents one fix attempt — review the history to understand what was tried.

How to Review

  • Each commit represents one fix attempt for plans/DA_Oauth_ts_Local_Debug.json
  • Compare the initial report with the final report to verify the fix
  • Check that coordinate/dhash changes are reasonable for the current VS Code layout

Changes Made to Test Plan

.../vscuse/vscode-test-cases/plans/DA_Oauth_ts_Local_Debug.json     | 6 +-----
 1 file changed, 1 insertion(+), 5 deletions(-)
View diff
diff --git a/packages/tests/vscuse/vscode-test-cases/plans/DA_Oauth_ts_Local_Debug.json b/packages/tests/vscuse/vscode-test-cases/plans/DA_Oauth_ts_Local_Debug.json
index 50a25a5..887fc4e 100644
--- a/packages/tests/vscuse/vscode-test-cases/plans/DA_Oauth_ts_Local_Debug.json
+++ b/packages/tests/vscuse/vscode-test-cases/plans/DA_Oauth_ts_Local_Debug.json
@@ -54,11 +54,7 @@
       "retry_count": 0,
       "continue_on_error": "0",
       "depends_on": [],
-      "preconditions": [
-        "dhash:348:92:16:5:a166aa18e6f5d5d5",
-        "dhash:348:92:96:5:204e21a0d926c357",
-        "dhash:348:92:0:10:c8c88094b2717075"
-      ],
+      "preconditions": [],
       "postconditions": [],
       "tags": [],
       "screenshot": "step_db5979ea",

Generated by vscuse-auto-fix.yml | model: claude-opus-4.8

…p_6f17d0b9 (click "M365 Agents Toolkit" activity-bar icon) in group__debug_in_copilot_local.json; the debug config dropdown is now a quickpick with an extra search box/C# entry that stays open and dims the background, so the icon-region dhash never matched → 30s precondition timeout. Clearing the preconditions lets the click execute (dismissing the dropdown); fresh dhashes are auto-recomputed after passing.
…step_f1471ec9 in group__debug_in_copilot_local.json) to the proven-working reference (629,484 + matching dhashes + delay:3 from step_1c0bfb73), replacing stale 635,487 coords/dhashes that caused the 30s precondition timeout.
…p_db5979ea ("Declarative Agent" click in the New Project quickpick); the new search box / extra C# entry changed the (348,92) region pixels so the perceptual hash never matched → 30s precondition timeout. Coordinate 348,92 is unchanged (matches sibling DA_Oauth plans).
…YPE step (step_0fec5113) in group__debug_in_copilot_local.json so the freshly-launched Chrome debug page has time to settle focus — without it the email never landed (field stayed empty), leaving Next at y≈516 and causing step_f1471ec9's y=484 filled-state precondition to time out; this mirrors the proven-working first-login type step (step_399e458d, which has delay:3).
…roup__debug_in_copilot_local.json from "${{var:app_name}}local displayed on the main section of the screen." to "${{var:app_name}}local can be found." — the app launched successfully (name visible in the Copilot Agents sidebar and message input placeholder) but the new M365 Copilot greeting UI no longer shows the app name in the main section, so the location-restricted wording failed; the new wording mirrors the passing remote-debug assertion.
…p_b4728564 ("Click the input box in the M365 Copilot chat page") in group__validation_DA_repair_Oauth.json — the 96px-wide dhash band at 376:409 covered the input placeholder "Message vscuse_app_*local" (randomized per-run app name) and the changed M365 Copilot greeting UI, so the hash could never match → 30s precondition timeout; coordinate (376,409) is unchanged and matches sibling validation groups, and fresh dhashes auto-recompute after the step passes.
…:512:384:0:20:11dc6346f8e6e4e8`) on step_22c29569 (type "show repair records assigned to karin blair") in group__validation_DA_repair_Oauth.json — the wide band centered at 512,384 covered the M365 Copilot main area, which now shows the new greeting "Hi, I can help you recap and prep for meetings" plus randomized per-run app name, so the hash could never match → 30s precondition timeout; the input box is confirmed present/ready (prior OCR click succeeded), and fresh dhash auto-recomputes after the step passes.
…_0115ed3f in group__debug_in_copilot_local.json) to the proven-working first-login click (step_be23e3b3: coords 369,350 + its three dhashes) — the Chrome sign-in page IS present in the failing screenshot, but step_0115ed3f's stale dhashes (389,351) never matched, causing the 180s precondition timeout; the sibling's hashes just matched the identical page this run, keeping the "wait for sign-in page" gate intact instead of removing it (which would fire the click before Chrome loads).
@github-actions

github-actions Bot commented Aug 1, 2026

Copy link
Copy Markdown
Contributor

❌ VscUse Test Plan — Tests failure

Why these tests: Selected 10 impacted test plan(s) changed in this PR (no product code modified).

Branch diff: auto-fix/DA_Oauth_ts_Local_Debug/2026-08-01dev

Results: ✅ 2 passed · ❌ 8 failed (of 10 plans)

Plans run:

  • DA_Api_Key_js_Local_Debug
  • DA_Api_Key_ts_Local_Debug
  • DA_Microsoft_Entra_js_Local_Debug
  • DA_Microsoft_Entra_ts_Local_Debug
  • DA_None_js_Local_Debug
  • DA_None_ts_Local_Debug
  • DA_Oauth_js_Local_Debug
  • DA_Oauth_js_Remote_Debug
  • DA_Oauth_ts_Local_Debug
  • DA_Oauth_ts_Remote_Debug

Step Status
1️⃣ Build VSIX (CD) ✅ Done
2️⃣ Build Docker image ✅ Done
3️⃣ Run UI tests ❌ Tests failure (2/10 passed)

🎯 Actual UI test run
🔗 Full pipeline results
📊 Detailed test report

ℹ️ How were these tests selected?

GitHub Copilot (Claude Sonnet 4.6, high reasoning) analysed the PR title, description, and the diff between
auto-fix/DA_Oauth_ts_Local_Debug/2026-08-01 and dev
to pick the most relevant test plans from packages/tests/vscuse/Index.md.

@github-actions

github-actions Bot commented Aug 1, 2026

Copy link
Copy Markdown
Contributor

E2E Test Selection — AI Selected

Why these tests: All changed files are under packages/tests/vscuse/** which maps to rule 8 (vscuse UI tests, not e2e tests) — no e2e test cases needed.

Cases selected (0):


View pipeline run

Need to run more tests?

Comment on this PR:

  • /e2e-run ./path/to/test.tests.ts — add specific cases to AI selection
  • /e2e-run-all — run all e2e cases

Then re-run the workflow.

@codecov

codecov Bot commented Aug 1, 2026

Copy link
Copy Markdown

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 87.87%. Comparing base (72fee7f) to head (1c617b1).

Additional details and impacted files

Impacted file tree graph

@@           Coverage Diff           @@
##              dev   #16498   +/-   ##
=======================================
  Coverage   87.87%   87.87%           
=======================================
  Files         648      648           
  Lines       34246    34246           
  Branches     8086     8086           
=======================================
+ Hits        30093    30095    +2     
+ Misses       2337     2335    -2     
  Partials     1816     1816           

see 2 files with indirect coverage changes

🚀 New features to boost your workflow:
  • ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
  • 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant