mobile-debug-mcp 0.31.0 → 0.32.1

package/docs/benchmarks/interaction-latency/final.md

@@ -0,0 +1,17 @@
+# Interaction latency benchmark
+
+Commit: d3e6ae720db1ad1da294795790689df5a80b56b0 (working tree candidate). Fixture: 1.0, SHA-256 98244deeeba0a21da7d99a3e4b43c97806d6f8a8787ad7fe38249e5d311a6fb6.
+
+Environment: Pixel_9_Pro, Android 16, API 36. Three warm-ups and ten measured runs per mode. Failed runs excluded.
+
+| Mode | p50 ms | p95 ms | Failures |
+|---|---:|---:|---:|
+| baseline | 124094.50754200004 | 124597.88916599983 | 0 |
+| full | 87262.29954100028 | 88067.6383750001 | 0 |
+| light | 51817.59008300025 | 53873.19891700009 | 0 |
+
+First failure by mode: baseline: none; full: none; light: none.
+
+Full p95 gate: true. Light p50 gate: true.
+
+See final.json for requests, timing components, and failure selectors.

package/docs/benchmarks/interaction-latency/ios-validation.md

@@ -0,0 +1,11 @@
+# Spec 018 iOS device validation
+
+Validated on 27 September 2026 with the booted iPhone Air simulator (iOS 26.4.1).
+
+- `npm run test:device` passed all six Android and iOS screenshot, log, and UI-tree smoke checks.
+- Over a spawned stdio MCP server, `start_app` launched Settings with `light`/`compact` and returned a fresh UI observation.
+- A selector-based `tap_element` opened General with `light`/`compact`; its result included an available state delta.
+- `run_journey` used two selector taps to return to General and open About, then asserted a visible `Model Name` label. All three steps passed with `light`/`compact`.
+- A separate `full`/`debug` selector tap returned to General and reported a fresh observation. A `none`/`compact` selector tap reopened About without requesting post-action verification; a separate UI-tree read confirmed the destination.
+
+The performance benchmark and its fixture are Android-only. These iOS checks exercised the interaction contract on a system app; they did not measure latency on an iOS app with network requests.

package/docs/benchmarks/ui-tree-performance/README.md

@@ -0,0 +1,51 @@
+# UI tree observation performance benchmark
+
+This paired Android benchmark compares a clean, built instrumentation baseline
+with a clean, built candidate. It runs the Spec 018 static journey and a local
+network-backed loading-to-ready flow against the same fixture APK and emulator.
+The network flow asserts a run-unique value from a physical tree read that
+started after the local server sent its response.
+
+Capture the instrumented baseline first from a clean, built baseline checkout:
+
+```sh
+UI_TREE_BASELINE_ROOT=/path/to/instrumented-baseline \
+LATENCY_DEVICE_ID=emulator-5554 \
+npx tsx test/device/automated/ui-tree-performance/benchmark.ts --baseline-only
+```
+
+This writes `baseline.json`. After measuring and implementing an optimization,
+run the paired comparison from the clean, built candidate checkout:
+
+```sh
+UI_TREE_BASELINE_ROOT=/path/to/instrumented-baseline \
+UI_TREE_CANDIDATE_ROOT=/path/to/optimized-candidate \
+LATENCY_DEVICE_ID=emulator-5554 \
+npx tsx test/device/automated/ui-tree-performance/benchmark.ts
+```
+
+Add `--smoke --baseline-only` for one attempt per flow while checking the
+fixture and emulator connection. Smoke mode writes no performance report and
+does not satisfy the sample-count or acceptance gates.
+
+Both input checkouts must contain `dist/server.js` and have clean Git working
+trees. The harness checks both before installing or resetting the fixture. It
+builds one APK from the candidate fixture and uses that exact APK in both modes.
+Each flow and mode receives three warm-ups and ten measured attempts; baseline
+and candidate order alternates by pair.
+
+The paired run writes `baseline.json`, `candidate.json`, `comparison.json`, and
+`comparison.md` here. Timing events are enabled only in the spawned benchmark
+servers through `MOBILE_DEBUG_MCP_UI_TREE_METRICS=1`; normal MCP responses do
+not gain timing fields. Events contain timings, platform, generated journey step
+IDs, and opaque acquisition IDs. They omit UI contents, app identifiers,
+screenshots, logs, payloads, and device serials. iOS latency remains unmeasured.
+
+The checked-in `initial-baseline.json` is the standalone capture taken before
+the optimization. The checked-in `baseline.json` is the baseline side of the
+paired run, so it shares the fixture APK hash with `candidate.json`.
+`comparison-attempt-1.json` retains an earlier paired run with two visible
+network failures; see `comparison.md` for the outcome and limitation.
+The report commit SHAs are available locally as `codex/spec-019-baseline` and
+`codex/spec-019-candidate`; create clean, built checkouts of those refs to
+reproduce the paired measurement.