# #4062 v4:native organic into radar: already there, no change (26 Oct, ppfeeds)
- Source: /opt/sw/radar/radar.jsonl on research-box (ev=cross, th 300k..20M; ev=cap5; fields venue, symbol, token, pool, launcher, cap_first, depth_usd, hooked, n_senders, fee, hook, lag_s, backfill).
- Radar covers every v4 PoolManager pool (any hook, WETH/native/USDG/SUSD quote). Heartbeat fresh (lag p50 1.8 s, head=block 98057483).
- Check, live rows since 24 Oct (backfill=false): GME 11 cross@300k v4:native (+1 hooked), BTC 8, LAUNCH 6, AMC 2 (+v4:usdg/v2/v3 pools for the same tickers). Symbols are filled, so a symbol-repeat gate can be replayed straight from radar.jsonl.
- Caveats for the replay: AMC row had depth_usd 0 and cap_first 677k (priced from a later swap, not an organic seed); keep thin/depth filters. Group by symbol across venues (same ticker appears as v4:native, v4:usdg, v2, v3).
- Example: python3 -c "import json;[print(d['symbol'],d['venue'],d['cap_first'],d['depth_usd']) for d in map(json.loads,open('/opt/sw/radar/radar.jsonl')) if d.get('ev')=='cross' and d.get('th')==300000 and d.get('venue')=='v4:native' and not d.get('backfill')]"
- Why mc5 did not see them: mc5 does not read radar v4:native (pickers skip v4); that is a strategy lane gap, not a feed gap.