Branch and PR evidence

R3/P3: settled state is correct, but the head briefly shows the old PR beside the optimistic new branch. Never-stale visual expectation fails.

T3 Browser panel, Chromium 157 on Linux; no device emulation; 1440×900 CSS pixels, DPR 2 (2880×1800 PNG/video), 100% browser/font scale, dark theme, UTC. Same sidebar/details widths and bottom scroll per row. Pointer parked on blank header at (1100,26) for stills.

Full evidence report, timestamps, masks, server projections and limitations · Verbatim upload output

R1/P1 — first feature → feature turn

R2/P2 — later turn → default branch guard

R3/P3 — real branch picker

Head R3 transient, playback 00:02.77

Scripts

• seed.sh
• launch.py
• rpc-driver.js
• project-driver.js
• scenario-driver.js
• switch.py
• transcript.py
• capture-endpoint.js
• picker-observer.js
• normalize.py
• make-diffs.sh
• make-sheets.sh
• videos.py
• freeze-clock.js (reference only, not executed)
• retake-r2.js (reference only, not executed)
• publish.py

Required rows captured: 6/6. See the report for supplemental rows not exercised. AI-assisted; verified with the tests and tools listed above.