# Wuji Follow-up Audit Summary: 4 Items

generated_at: 2026-06-14T19:05:40+08:00

## 1. Public 5m Backfill For Three-day-high Sell
- rows: 132
- fetch_status_counts: {"FETCH_OK":132}
- audit_status_counts: {"PASS_5M_PROXY":132}
- conclusion: all missing 1m SELL_THREE_DAY_HIGH_NOT_RISING rows now have public 5m proxy evidence and pass the directional check.

## 2. 5m Proxy For Non-exact Minute Events
- input_gap_events: 1090
- gap_events_with_time: 1078
- unique_symbol_dates_requested: 695
- fetch_status_counts: {"FETCH_OK":695}
- proxy_status_counts: {"PROXY_5M_BAR_FOUND":1078,"NO_EVENT_TIME":12}
- conclusion: all 1078 gap events with event_time are mapped to public 5m proxy bars; 12 rows have no event_time and cannot be mapped without rule-specific timing.

## 3. SELL_OPEN_VOLUME_STALL Independent Scan
- lot_rows: 735
- coverage_rows: 790
- coverage_status_counts: {"SCANNED_PROXY_5M":370,"SCANNED_EXACT_1M":420}
- candidate_rows: 0
- conclusion: no independent EXACT_1M or PROXY_5M candidates were found under open_volume_ratio >= 10 and 1%-2% first gain.

## 4. Intraday Breadth Proxy
- target_dates: 251
- minute_files_scanned: 2429
- output_rows: 3514
- total_observations: 7951075
- eligible_count_range_all_times: 1969-2406
- eligible_count_range_1040: 1969-2406
- conclusion: generated local non-empty minute-file pool breadth; this is not full A-share market breadth because local minute coverage is incomplete.

## Main Artifacts
- public_5m_proxy_summary.md
- minute_event_5m_proxy_summary.md
- minute_event_5m_proxy_audit.csv
- open_volume_stall_scan_summary.md
- open_volume_stall_independent_candidates.csv
- intraday_breadth_minute_pool_summary.md
- intraday_breadth_minute_pool.csv