[Drift Audit] Maestro rule doc lists stale test category directories

[daxmobile]

Task/Issue URL: (app.asana.com/redacted)

Description

.cursor/rules/maestro-ui-tests.mdc — the "Types of UI tests" section described a .maestro/ layout that no longer exists. The doc is the map agents use to locate and run Maestro suites, so a wrong directory name produces a maestro test command that fails.

Verified against the .maestro/ tree on develop (ee431b8):

Directories listed under names that no longer exist — the doc's commands (maestro test .maestro/sync) do not resolve:

Doc said	Actual directory
ad_click_detection	ad_click_detection_flows
ads_preview	ads_preview_flows
sync	sync_flows

preonboarding listed as a top-level category, but no such directory exists. Those tests now live inside .maestro/onboarding/ as preonboarding_returning_user_quick_setup_customize.yaml and preonboarding_returning_user_quick_setup_start_browsing.yaml. The doc now folds pre-onboarding into the onboarding entry.

Directories absent from the list. The most significant is unified_input_screen: #9367 (merged 2026-08-03) turned its unifiedInputTest tag back on (tags: [] → tags: [unifiedInputTest] in unified_input_open_duckai_back_to_ntp.yaml) and un-commented the AI-off omnibar-position combos, making it an active tagged CI suite. An agent reading the doc would conclude input-bar tests live only in input_screen. Also added, each described from its own flow headers: binary_upload, deeplink, omnibar, performance, privacy_tests_internal, serp.

The section keeps its "(non-exhaustive)" framing; only the category list changed. Every entry in the updated list resolves to a directory on disk, and every directory in .maestro/ other than shared/ is now listed. No test files, tags, or code were touched.

Note on scope: these are bare backticked names, not paths, so AiConfigChecker.extractReferences never classified them as references (it only treats tokens containing / as paths, or :-prefixed tokens as modules) — this drift was invisible to aiConfigCheck rather than a regression in it.

I also re-checked .cursor/rules/wide-events.mdc against the sampling work in #7965 / #9346 and .cursor/rules/pixel-definitions.mdc against #9279. Both still match the code — samplingProbability, the SAMPLED_OUT_FLOW_ID sentinel, FlowStatus, CleanupPolicy, the pixel/POST transport split and its default toggle values, and the PixelDefinitions npm scripts all read correctly. No changes needed there.

🤖 AI-docs Drift Auditor

Steps to test this PR

• Read the updated rule doc against the linked code change and confirm it now matches.
• Confirm every backticked category name in the list is a real directory: for d in $(grep -oP '^- \K[a-z0-9_]+(?=)' .cursor/rules/maestro-ui-tests.mdc); do [ -d ".maestro/$d" ] || echo "MISSING: $d"; done — prints nothing.
• Confirm nothing is left out: for d in .maestro/*/; do n=$(basename "$d"); [ "$n" = shared ] && continue; grep -q "^- \$n`" .cursor/rules/maestro-ui-tests.mdc || echo "UNLISTED: $n"; done` — prints nothing.
• Confirm .maestro/unified_input_screen/ carries the unifiedInputTest tag, and that pre-onboarding flows are the preonboarding_*.yaml files inside .maestro/onboarding/.

UI changes

Before	After
No UI changes	No UI changes

Note

🔒 Integrity filter blocked 7 items

The following items were blocked because they don't meet the GitHub integrity level.

• #8891 search_pull_requests: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #9341 list_pull_requests: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #8913 search_pull_requests: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #8899 search_pull_requests: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #8777 search_pull_requests: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #8487 search_pull_requests: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".
• #8381 search_pull_requests: has lower integrity than agent requires. The agent cannot read data with integrity below "approved".

To allow these resources, lower min-integrity in your GitHub frontmatter:

tools:
  github:
    min-integrity: approved  # merged | approved | unapproved | none

Generated by AI-docs Semantic Drift Audit · ◷