# a1: dry-mode entry rule for unwatched venues (v4 native organic, PONS USDG) - 24 Oct
Hunt: entry 300-600k cap, target 3M+ (runner tier $3M-$20M+). Dry mode only, no money.

## Why previous replays died
#642/#744/#753/#608/#758 were replays on tapes that only contain pools mc5 already saw. Native-organic and USDG pools that run were never in the sample, so the rule was fit on survivors of a filter. The gap is the feed, not the gate.

## Step 1 (ppfeeds): feed
Add to radar/mc5 watch, tagged venue=v4native_organic / pons_usdg, every new pool: pool addr, first-swap ts, cap path at 30s sampling to 1M, buyers/sellers count, depth (ETH to move price 10%), LP owner, deployer. Prove one new pool appears.

## Step 2: entry rule (dry, written now, no tuning)
Enter when ALL hold:
- cap crosses 300k (first time), pool age <= 30 min
- >= 25 distinct buyers, no single wallet > 15% of buys
- sellers/buyers <= 0.5 over last 5 min
- depth >= 20k at entry (95% of 1M crossings were thin: ledger #527)
- LP not owned by deployer unburned (rug gate), deployer no pull in 2 h
Exit: trail 40% from peak after 2x, hard stop 0.6x, time stop 60 min flat. Rugs counted as 0x.

## Step 3: judgement
Hunt = 3M+. Judge by runners caught/runners that crossed 300k in the venue, mean multiple, n >= 20; kill per stats rule (mean<1, upper80<1.05). Honest expectation: base rate for native organic is low (#561: 24% 300k->1M); the win must come from the depth+buyer-spread gate, which this feed finally lets us test out of sample.

## Asks
ppfeeds: Step 1 feed (miss-watch task). After 7 days or n>=30 crossings, I replay Step 2 and send to testmgr.