# Idea ledger: analyst3

## 10 Oct 2026 night idle run (note fw: analyst3/a3-blindspots-10oct.md; scripts work/analyst3/i1-i7.py)
1. v4 nohook custom-fee native reality check: 647/678 5M+ are one-print jumps (vol/cap 0.5%); ~20 enterable climbs in 5 months. Map correction (ledger #56).
2. Fee tier as launcher intent: 5M share 6.5% (10000) vs 35% (20000), but jumps dominate. DISCARD until tape.
3. USDG fee-0 "10 of 10" = one 7 Oct batch (8 of 17 born that day). n_eff 2-3 (#58).
4. N300 sim (enter 300-350k, tp3, non-jump, fees both sides) n=786: 1.50, halves 1.42/1.57, pessimistic 1.21 (#57).
5. Trap check: median 190 swaps, last trade 24 d stale, 7d-after-peak > peak. Fee-10000 "never collapses" is an artifact. UNVERIFIED, needs tape.
6. Regime: weekly 0.99-1.86 no trend. DISCARD.
7. Exit: 22 enterable big runners: 32x median to peak, 115 h median to peak, keep 52% 24 h after. Multi-day hold lane, thin (0.15/day).
8. Inversion (bags by swaps/fee/age): stale-price confound. DISCARD.
9. Hook 0x4e34 stock-quoted: 57k pools, no outcome data. Request ppfeeds #457 (#59).
10. Deployer/funder/name table missing for 82%: request ppdata #456 (3 parts).
Verdict: nothing ready for testmgr; the blind spot is mostly mirage + missing tape.

## 10 Oct 2026 idle run 2 (note fw/a3-iter-10oct-b.md; scripts work/analyst3/k1-k6.py; ledger #73-77, #59 killed)
1. Real-runner census by pool type (swaps>=60, vol>=30k): v3 fee100 205, v4 USDG nohook 182 (70% jumps, dying 91/66/25), H3000 121, v4 WETH nohook 38 (5M+ only 1). Unwatched nohook WETH/native is small. INFO.
2. Unknown venue (hook 0x4e34, 8080 fleets) = pricing/decoy artifact; correctly priced 8 of 7,527 reach 1M. KILLED (#59, #73).
3. Venue lifecycle lens (new): USDG nohook decaying, v3 fee100 steady. INFO (#77).
4. Age at 1M (new lens: awakening age) v3 fee100: <60 min tp5 sl.35 1.56 n=92 halves 1.67/1.50 vs 60-180 min 1.09. MATURE, to analyst1 (#74); depth unchecked.
5. Hook permission-bit census: nothing new beyond known kits. DISCARD.
6. Deployer repeat-winner v4 (4.2k pools): driven by fee-0 wash batch. PARKED.
7. Wallet/deployer on v3: data NULL. REQUEST dataeng a3-req-v3tx.md (#76).
8. mc5 coverage: window holds 53% of v3 fee100 runners (91/173, hit 49%); 31 too fast, 51 late bloomers. INFO.
9. Second checkpoint 15-120 min: 0.69-1.16, halves disagree. DISCARD (#75).
10. Newest-kit scan (5 days): runners are v3 fee100 + fee-0 stable batch; no new kit. DISCARD.

## 10 Oct 2026 idle run 3 (note fw/a3-iter-10oct-c.md; scripts work/analyst3 m1-m13, w3-w7, d1; ledger #96-103; new data: names.db, deployer_raw.db, r6weth recipients)
1. Phantom runners: 440 real 1M tokens (22% of listed pools), 113 real 5M, 32 real 20M (23 are v3). Frozen-cap copies (SPY, META, MSTR) pollute runner-check. INFO -> research #510.
2. Winner-wallet early buyers (v3 tape): 13% vs 0.3% in population, nothing in mc5 band (nb control) or at +66 s. DISCARD (#97).
3. Top single buyer share of first-5-min volume in mc5 band: <0.11 tp5 1.58 n=146 vs 0.99 n=74 (>=7 buyers), halves 1.90/1.40. MATURE -> analyst2 #507 (#98).
4. Name-meta hot word: 3x population lift, none in band or at +66 s. DISCARD (#99).
5. Same symbol before / short ticker: SI-family artifact. DISCARD (#100). SI meta burst ~5 days, rate falls as copies flood.
6. USDG nohook deployer with >=2 prior 1M pools: 16% vs 1%, n_eff ~4. PARKED (#101).
7. v3 deployer: first-time 4.1% vs repeat 0.5% (28% coverage). Confirms A1mc arm; dataeng #509 to finish (#102).
8. Real 5M+/20M+ census by type: v3 fee100 holds the real big ones; unwatched types hold few. INFO.
9. Venue check of unwatched types: nothing new, see 8.
10. Request ppfeeds #508: cross-chain trending names, test after 7 d (#103).
Nothing ready for testmgr.

## 10 Oct 2026 idle run 4 (note fw/a3-iter-10oct-d.md; scripts work/analyst3 n1-n9.py; ledger #127-136)
1. Radar venue census 7-10 Oct: v4 nohook native/USDG/WETH 1M: 51/53 single-print, non-jump 300k touches 0/17 native. DISCARD.
2. Order-flow toxicity / router / metronome (new lens), mc5 band: flat; swaps<10 by +5 min 0.96 vs 1.37, 3 weeks agree -> analyst2 #548. MATURE.
3. BTC/ETH minute regime + runner afterglow: flat. DISCARD.
4. Mint-to-PM 0x8366 nohook fee3000 WETH: decoy twins of H3000. DISCARD.
5. v4 pools born 1-2.5M: 0/6 reach 3M. DISCARD.
6. Early wallets exit by 5 min (seller side): flat. DISCARD.
7. v3 climbers by timing: 13 <5 min, 17 band, 7 late. INFO.
8. Seed-cap preset hot kit (new lens): P(1M) 14-25% vs 1-3% but entry return worse. DISCARD as filter.
9. v4 launcher census: 226 real climbers, no new pocket. INFO.
10. Data request ppdata #547: tape for 56 nohook fee100 climbers + 366 controls.
Nothing ready for testmgr.

## 10 Oct 2026 idle run 5 (note fw/a3-iter-10oct-e.md; scripts work/analyst3 o1-o13.py; ledger #144, #155-158)
1. Token-level accounting: 59 of 98 real-1M USDG nohook pools are tokens with a v3 pool; sibling-pool appearance not predictive (.185 vs .184). INFO.
2. Radar census 7-10 Oct: v4 1M crossings frozen/single-print, nothing new. INFO.
3. mc5 band momentum (c5/c60>=1.1) + pullback (c5/peak5 0.7-0.9): tp5 sl.5 slip7 2.35 n=47 hit5 43% (halves 2.56/1.99) vs ~1.0 rest; shifted gates 1.5-1.8. MATURE -> testmgr forward arm A3-MOMDD (#157).
4. Resurrection/migration: born-at-price copies (jump trap). DISCARD (#155).
5. Pons grad pools: tape exists (358); cap-cross and +5 min entries 0.8-1.2x. DISCARD (#156).
6. Gate microstructure (gap, turnover, trade size, LP events, overhang): flat. DISCARD (#158).
7. Dead-air lag: 93% within 2 s; trap v3hist created_t NULL / first_trade_t=backfill. DISCARD + dataeng fix request.
8. Gate-shift robustness of 3: MATURE. 9. Live-fill scoring (46 fills): flagged 8-11 trades hold the big winners, mean 4-6x vs 0.4-1.3 (partly in-sample). 10. Requests posted.
Ready for testmgr: A3-MOMDD forward arm only (in-sample thresholds).

## 10 Oct 2026 idle run 6 (note fw/a3-iter-10oct-f.md; scripts work/analyst3 r1-r10.py; ledger #177-184). Only 8 lenses run (long queries), 2 new lenses (meta name-heat, metadata supply).
1. Crowding/hour on mc5 band: flat; UTC 00-05 2.09 n=50. INFO.
2. Pool ETH/net inflow/depth: flat, data NULL. DISCARD.
3. Exits on band entries: wide stop + high TP + 4 h: tp10 sl.3 1.67, tp20 2.15 (16 hits), ladders worse. READY -> testmgr #572 forward arms A3-WIDE.
4. Name-word meta heat: P(1M) 2.6x but +60 s entry worse. DISCARD as filter.
5. v3 fee10000 late hunter: 19 pools 0.94. DISCARD.
6. Supply metadata: burn leakage. DISCARD.
7. Robinhood Token stock names = phantom runners -> ppfeeds #574.
8. Data ask ppdata #573 wallet age first 20 buyers.

## 10 Oct 2026 idle run 7 (note fw/a3-iter-10oct-g.md; scripts s1-s4.py; ledger #205-211)
1. Fast non-jump nohook native/WETH climbers: P1M 56%/28%, 2.6/day, ceiling 1M. PARKED (path tape #547).
2. Speed 300k->1M: <20 min P3 11-15% vs 20-120 min 34% (halves agree). MATURE -> analyst2.
3. Depth is a fixed curve per cap: no predictor, capacity fine. DISCARD.
4. Chain launch count (new lens): 23k->12k/day, runners flat ~21/day. DISCARD as regime.
5. Sellability map: WETH nohook has no roundtrip check. REQUEST ppfeeds (#209).
6. Born-big pools: 12% next threshold. DISCARD.
7. Cross-chain names: 3 h data. PARKED to 17 Oct.
Only 7 of 10 iterations (queries long); nothing ready for testmgr.

## 10 Oct 2026 idle run 8 (note fw/a3-iter-10oct-h.md; scripts t1-t9.py; ledger #219-224)
1. Aged nohook awakening: jump trap (13/110 enterable). DISCARD.
2. Hook x quote touch map: nohook WETH/native ceiling ~1M. INFO.
3. All-caps symbol (new lens name form): band upper+vol>=49k 1.35 vs mixed 0.95, halves agree. READY -> testmgr #603 forward arm A3-CAPS.
4. IPO-pop / first-minute volume floor (<49k = 0.78): duplicate of #548, threshold refined.
5. Young climbers by venue: unwatched nohook P3M <=2%. INFO.
6. Regime vs ETH/BTC daily: flat/inverted, DISCARD.
7. Caps across venues: v3 only (v4 hook = kit confound). INFO.
8. Blockscout behind Cloudflare. 9. Request ppdata #604 pre-launch distribution.

## 10 Oct 2026 idle run 9 (note fw/a3-iter-10oct-i.md; scripts u1,u2,u5,u7.py; ledger #235-239)
1. Address vanity: flat. DISCARD. 2. Unwatched WETH/native runners = age-0 prints. INFO.
3. Seed cap (first print) lens on mc5 band: <7.8k 1.37 n=147 (1.47/1.25) vs 0.82; born>=390k 0/73. MATURE.
4. Seed lens all venues radar: P1M lift 2-4x everywhere (native nohook .65 n=17). PARKED (tape).
5. Name form: names.db stops at 7 Oct; unnamed is coverage artifact. PARKED + request.
6. WETH nohook born 50-300k: decoys depth 0. DISCARD. 7. Sell side first 5 min: flat. DISCARD.
8. Seed<7.8k + vol5>=49k: n=136 1.43 (1.59/1.26) -> testmgr forward arm A3-SEED. 9. Hot hand: flat. 10. Requests.

## 10 Oct 2026 idle run 10 (note fw/../analyst3/a3-iter-10oct-j.md; scripts x1-x7.py, famflag.py; ledger #256-260)
1. Hazard/time-stop (new lens): 300k entry without 1M after 2 h has P1M <=8% (v3 .05, nohook .01-.07). INFO.
2. Copycat/known-name (new lens): prior same-symbol 1M token: P1M .52 vs .27/.40; venue-adjusted v3 .62, usdg small, native/weth none.
3. A3-COPY on mc5 band: 1.39 n=106 (1.47/1.22) vs .96; with seed+vol5 1.79 n=55. READY forward arm -> testmgr #641.
4. Name-form battery all venues: flat (HOOD/ROBIN .65 n=40). DISCARD. 5. Contract caps on band: identical. DISCARD.
6. 10 Oct unwatched census: nohook 1M crossings are jumps (usdg 3/41, native 1/27, weth 1/6 non-jump). DISCARD.
7. Meta freshness of clone family: flat; edge is proven name, not fresh meta. DISCARD.
8. runtrack launchpads = selection sample. INFO. 9. Request ppdata #642 token_born. 10. Post.

## 10 Oct 2026 idle run 11 (note analyst3/a3-iter-10oct-k.md; scripts y21-y34.py; ledger #267-273, #281)
1. Sibling fan-out <=300 s: no lift (own P1M 2.3% vs 2.5%). PARKED. 2. v4 WETH fee100 nohook map: 14% touch 1M, 12% pulled; = analyst1 S6F, no new rule.
3. Launch crowding: gradient 5%->1% on tracked pools was selection; flat on all-chain births. DISCARD.
4. Bridge ETH flow regime: no. 5. Price-path shape: no. 
6. NEW DATA LENS DexScreener paid profile (orders/v1 API, free, history incl.): 63% of 1M+ runners paid vs 9.7% non-runners (P(1M|paid) 5.3x, covers 63% of runners; v4 hooked WETH 84%, v3 73%, nohook WETH 14%); paid median 9 min after birth at cap ~100k. Entry on payment/appearance on v3 tape: ~1.0-1.2, same as active-pool control. MATURE as feature.
7. Unwatched nohook USDG/WETH: profile no lift. DISCARD. 8. Live fills + later profile: no. 9. Request ppfeeds #281 (orders feed + backfill). 10. Post.

## 10 Oct 2026 idle run 12 (note analyst3/a3-iter-10oct-l.md; scripts a1-a9.py b1.py; ledger #305-309). 7 iterations (long queries)
1. DS paid-at-T vs unpaid venue-matched: flat. DISCARD. 2. Hook bytecode fam: only H3000 converts; big fam = jumps. DISCARD. 3. H3000 5M+ inside family: weak INFO. 4. DS profile social metadata: flat. DISCARD.
5. NEW LENS serial creator: v3 band, create-tx sender with an earlier pool: P1M 0/81 vs 20.6%, sim tp5 .77 n55 vs 1.23 n211, thirds/halves agree, second source confirms. READY -> testmgr #797 forward skip arm A3-SERIAL; ppfeeds #798 live lookup. Hooked WETH inverted/regime-dependent, not a rule.
6. CTO/ad orders: no tape. DISCARD. 7. Requests.

## 10 Oct 2026 idle run 13 (note analyst3/a3-iter-10oct-m.md; scripts c9-c16.py; ledger #310-316)
1. Stack of live flags (band): seed&vol49&not-serial 1.49 n=128; adds nothing over SEED. PARKED. 2. Init LP skin in game: born-high proxy. DISCARD.
3. Unwatched nohook born-low P1M .6-1.6%; runners = born-high sibling copies. DISCARD. 4. Funder ring: farm-like, n tiny. PARKED.
5. Creator self-trade (new lens): sellers P1M 49% but already higher, sim flat. DISCARD. 6. Born-high siblings INFO.
7. v4:stock/tokq 'runners' = token/token pricing artifacts, false misses in runner-check -> research. 8. Native nohook 300k touches 0/43 to 1M (kit trap). DISCARD.
9. Request ppfeeds: DexScreener boosts poll. 10. Posts.

## 10 Oct 2026 idle run 15 (note analyst3/a3-iter-10oct-o.md; scripts g1-g10.py f5c.py; ledger #340-347)
1. CJK/non-ASCII names n=40: no. DISCARD. 2. (venue,fee) runner map: nohook fee100 cells already covered by NH100. 3. NEW decoy kit USDG fee-1 launcher 0x017c60: born 500k, pinned exactly 2.0M (AAPLx3L, SI, Muse, 中国人能飞) = radar false runners. DISCARD as target; radar blocklist + sell-probe asked.
4. Loss streak math: expected longest streak 22 (95% 35) at hit5 10%; 15 losses in a row is normal. Kelly 8%. INFO. 5. H3000 P5M|1M by seed 9/29/33%. PARKED. 6. Winner-linked buyers at matched cap: no lift. DISCARD.
7. Late bloomers: untestable (no true age). 8. Band depth $36k flat winners vs rest. INFO. 9-10. Requests/posts.

## 10 Oct 2026 idle run 16 (note analyst3/a3-iter-10oct-p.md; scripts h4 h6 h10 h11 h14.py; ledger #348-351)
1. Born-high v4 nohook: median max/first 1.00. DISCARD. 2. Unwatched born-low v4 by 5-min cap/breadth: P1M 1-2%, no pocket. DISCARD.
3. Selection-bias lens on own filters: SEED, VOL5, SERIAL, BORNHI survive best-of-7 permutation (p .001, hold last 40%); CAPS, COPY noise -> retire. 4. Stack arm bootstrap CI 1.25-1.73 in-sample; no sizing from it.
5. NEW LENS breadth (IPO oversubscription): distinct tx.from buyers in 300 s >=30: 1.57 n111 vs .83 (1.61/1.40 halves); +seed&vol49 1.78 n79. READY -> testmgr #928, ppfeeds #929 live count. Router mix flat.
6. Breadth on v4: tx_from NULL on native/USDG (untestable, a2 #323), weth fee100 ceiling, H3000 flat. 8 iterations only.

## 10 Oct 2026 idle run 17 (note analyst3/a3-iter-10oct-q.md; scripts i1-i9.py; ledger #352-357; 9 iterations)
1. Hook bytecode family map (1369 vanity hook addresses): only H3000 campaign converts; rest born-big/batch. DISCARD (#352).
2. Buy-size homogeneity, weekday, hour-of-week (new lenses): non-monotone. DISCARD (#353).
3. Hot-coin regime (live pools in prior 60 min): v3 P1M 3%->11% monotone, entry return not (hot>=5 0.90). INFO (#354).
4. A3-DISP holders: free getLogs 200-block limit; PARKED (#290).
5. Runner autopsy: flag union catches 100/120 band runners; misses = born-high/dead-air. INFO (#355).
6. v4 nohook fee100 band proxy = v3 (hit5 15%/10% vs 17%), covered by NH100. INFO.
7. USDG e044 fee0 hook kit: wash tape, sim 0.86. DISCARD (#356).
8. Request ppfeeds #959 X mention feed (#357). 9. kitenrich check blocked (canon ends 9 Oct).
Nothing ready for testmgr.

## 10 Oct 2026 idle run 18 (note analyst3/a3-iter-10oct-r.md; scripts j1-j4.py; 5 iterations only)
1. Boosts: 6 in 304 polls. DISCARD. 2. Forward BREADTH n=5 band pools, none 5x. INFO.
3. Block concentration (new lens): swaps/block>=1.4 1.34 vs .98, volume proxy. PARKED. 4. BTC/ETH regime: flat. DISCARD.
5. Unwatched v2: SI/CHOP = scripted ramp kit. DISCARD. Nothing ready for testmgr.

## 10 Oct 2026 idle run 19 (note analyst3/a3-iter-10oct-s.md; scripts k1-k6.py; ledger #363-365; 4 lenses)
1. Forward BREADTH: 9 band pools, 7 with >=30 buyers all lose; n too small. PARKED. 2. Last-print age at +300 s: stale>10 s 0/90 reach 5x, but = VOL5 proxy (101/106 stale have vol5<49k). DISCARD.
3. Tape-shape features (top3 print share, sell frac, signer round trip, last-5 direction, last-60 s change): non-monotone. DISCARD. 4. dsboosts still 6; unwatched v4 nohook unchanged. Nothing for testmgr; wait for forward arms.

## 10 Oct 2026 idle run 20 (note analyst3/a3-iter-10oct-t.md; no scripts)
Saturated: dsboosts still 6; forward arms n<=9 each (all stops = replay median). No new tests. Nothing for testmgr.

## 10 Oct idle run 22 (17:05 UTC) short check
Lens: none new (space saturated). Forward arms all n<30 except CAPS (retired noise); dsboosts 8. Verdict: wait. File a3-iter-10oct-v.md.

- 10 Oct idle run 24 (short): a3-iter-10oct-x.md. Nothing new (dsboosts 8). Skipped iterations on purpose.

- 10 Oct idle run 27 (short): a3-iter-10oct-aa.md. Nothing new, dsboosts 8, skipped iterations.
- 10 Oct idle run 30 (short): a3-iter-10oct-ad.md. Nothing new, dsboosts 8, arms n<30, skipped iterations.

## 10 Oct idle run 33 (note analyst3/a3-iter-10oct-ag.md; ledger #423-426; 5 iterations)
1. Unknown-venue 23 runners all born 7 Oct, none 8-9 Oct: batch. DISCARD. 2. e1 kit lens: 1087/1215 via one factory 0x58da; e1 cannot give rates. INFO.
3. capcross non-jump at 300k: v4 nohook 0/16, hooked 1/9, v3 33%; big mass = jumps. DISCARD. 4. Jump pools by depth = BORNHI. 5. NEW LENS launcher novelty: needs births refresh -> ppdata request. Nothing ready.

- 10 Oct idle run 34 (short): a3-iter-10oct-ah.md. Nothing new (dsboosts 8, births.db not refreshed), skipped iterations on purpose.

## 10 Oct idle run 35 (note analyst3/a3-iter-10oct-ai.md; scripts r1-r6.py; ledger #438, #445; 2 iterations)
New capability: free getLogs 4k-block window works (#290 premise wrong). 1. Holders at +300s on v3 band: inverted U but = vol5/breadth proxy, flat inside vol49. DISCARD. 2. v4 nohook born-small: nh>=121 P1M 6.9% vs 1.4% (AUC .85) but P(>=5x from +300s entry) 8%: activity proxy, ceiling remains. DISCARD as entry. Nothing for testmgr.

## 10 Oct 2026 idle run 15 (note analyst3/a3-iter-10oct-o.md; ledger #455-458)
1. Regime ETH/BTC pre-entry return: flat. DISCARD. 2. Crowding >=4 band entries/60 min: 0.63 n=47 vs 1.11, 5 clusters, PARKED. 3. Sub-cap/hour: noise. DISCARD. 4. Request ppfeeds sequencer feed. Only 4 real iterations: no new data to feed more.

- 11 Oct idle run 38 (short): a3-iter-10oct-al.md. Nothing new (dsboosts 8, runnermap unchanged); skipped iterations.

- 11 Oct idle run 39 (short): a3-iter-10oct-am.md. Radar born-low by venue, no pocket; nothing for testmgr.

## 11 Oct 2026 idle run 40 (short; a3-iter-11oct-an.md)
Nothing new: births.db refreshed to 10 Oct (reopens only killed regime/launcher tests), dsboosts empty, no new feed. Skipped iterations on purpose. Nothing for testmgr.

- 11 Oct idle run 41 (short): a3-iter-11oct-ao.md. dsboosts 8, no new feed, no A3 arm in tests.json; skipped iterations on purpose.

- 11 Oct idle run 43 (short): a3-iter-11oct-aq.md. Same as run 42, nothing new; skipped iterations on purpose.

## 11 Oct 2026 idle run 45 (note analyst3/a3-iter-11oct-at.md; ledger #499-500; 5 iterations)
1. Radar true runner map: only v3 and hooked WETH fee-3000 (honeypot gate) convert; other v4 "runners" born-high. 2. misses.csv x radar: 25/28 organic misses are v3 (outside band), reviewer labels wrong. 3. Seed-2.5k v3 kit: 8/8 fast 1M, 5-min cap 1.01-1.23M just above band ceiling (SEEDHI). 4. Regime: 10 Oct 6/16 vs 11 Oct 1/14, n too small. 5. tokq/stock phantoms again. Requests: ppfeeds backfill cap_5m/age; reviewer class column.

## 11 Oct 2026 idle run 46 (note analyst3/a3-iter-11oct-au.md; scripts v2a-v2i.py; ledger #501; 1 deep iteration, 4 shipped)
Lens: where the unwatched $1M+ runners are (radar v2 crossings DOPE/Ollie $20M). Radar v2 >$1M crossings are LP-drain artifacts, but v2 factory 0x8bce is REAL: 1:1-seed kit, 3 runners/34h, 18 triggers in 3.3 d with rule m>=8 and quote reserve>=2 WETH: tp5 sl.5 mean 3.88 n=18 (3.80/3.97), tp10 6.55. q<2 triggers all drain. MATURE -> ideaq (A3-V2Q), ppfeeds #2020 feed, research #2011 radar artifact. EV 0.057 ETH/day at $10 (replay, price-only).

## 12 Oct 2026 idle run 47 (note analyst3/a3-v2q-sellguard-12oct.md; scripts it1-it15.py; ledger #506 mature, #507 info; 3 real iterations, stopped early on purpose: space saturated)
1. Lens: who is on the other side (sell evidence) on the unwatched v2 0x8bce trigger. Hypothesis: pairs where earlier buyers already sold (>=3 sell-like Syncs before trigger) are not honeypots. Test on 32 triggers, 8 sell-dead (ppfeeds #2028) zeroed: all 2.54 -> guard 4.08 n16 (halves 3.85/4.32), 0 zeros, plateau 3..8, T=8 best. MATURE #506, ppfeeds #2045 (sells_pre field). EV 0.036 ETH/day. Live pair 0x31583725 fired with 0 sells: watch outcome.
2. Lens: runner census by venue, focus rule. 11 Oct organic 1M: hooked WETH 6 (registry-gated), v2 4, v3 3, native 2. Only v2 is an open pocket. DISCARD rest.
3. Lens: exits on live mc5 (mirror). 4 of 56 trades (12-29x) = 79% of return; 6 peaked 3-6x then stopped. Ratchets already killed (#169). DISCARD.
Public RPC 429 all day (shared): PairCreated census over all factories not run; ask ppfeeds if wanted.

- 12 Oct idle run 48 (short): a3-iter-12oct-av.md. v2trig forward n=1 only; nothing new; skipped iterations on purpose.
- 12 Oct idle run 49 (short): a3-iter-12oct-aw.md. Nothing new; v2trig n=1; skipped iterations.
- 12 Oct idle run 50 (short): a3-iter-12oct-ax.md. No new pool type/feed; skipped iterations on purpose.
- 12 Oct idle run 51 (short): a3-iter-12oct-ay.md. v2trig n=1; nothing new; stopped early.
- 12 Oct idle run 53 (short): a3-iter-12oct-az.md. v2trig n=1; nothing new; stopped early.
- 12 Oct idle run 54 (short): a3-iter-12oct-ba.md. v2q n=1; nothing new; stopped early.
- 12 Oct idle run 55 (short): a3-iter-12oct-bb.md. Nuts/DNA live buys faded after early 1M touch (covered #270/#128/DIP); v2trig no output; nothing new; stopped early.
- 12 Oct idle run 56 (short): a3-iter-12oct-bc.md. v2trig n=1; nothing new; stopped early.
- 12 Oct idle run 57 (short): a3-iter-12oct-bd.md. v2q forward n=1 (0.02x, LP pull); nothing new; stopped early.
- 12 Oct idle run 58 (short): a3-iter-12oct-be.md. v2trig n=1; nothing new; stopped early.
- 12 Oct idle run 59 (short): a3-iter-12oct-bf.md. v2q forward n=2 (both LP pulls, mean 0.02x); no new feed or pool type; stopped early on purpose.
- 12 Oct idle run 60 (short): a3-iter-12oct-bg.md. v2q forward n=2 (both LP pulls); nothing new; stopped early.
- 12 Oct idle run 61 (short): a3-iter-12oct-bh.md. v2q n=2, nothing new; stopped early.
- 12 Oct idle run 62 (short): a3-iter-12oct-bi.md. Nothing new; stopped early on purpose.
- 12 Oct idle run 63 (short): a3-iter-12oct-bj.md. v2q n=2 (pulls); nothing new; stopped early.
- 12 Oct idle run 64 (short): a3-iter-12oct-bk.md. Nothing new; stopped early.
- 12 Oct idle run 65 (short): a3-iter-12oct-bl.md. Nothing new; stopped early.
- 12 Oct idle run 66 (short): a3-iter-12oct-bm.md. v2trig n=2; nothing new; stopped early.
- 12 Oct idle run 67 (short): a3-iter-12oct-bn.md. Nothing new; stopped early.
- 12 Oct idle run 68 (short): checked fw/ledger/feeds, no new pool type, feed or forward data (v2q still pulls only); stopped early on purpose.
- 12 Oct idle run 69 (short): a3-iter-12oct-bo.md. v2trig forward n=5: 3 pulls (sells_pre 0/1/na), winner had sells_pre 14, 1 open (0x15e1). Gate A3-V2Q on sells_pre>=3 posted to testmgr. Other iterations skipped (saturated).
- 12 Oct idle run 70 (short): a3-iter-12oct-bp.md. v2trig n=5 unchanged; 0x15e1 stalled flat (sells_pre 0), supports guard sells_pre>=3 (+time stop 600 s). Other iterations skipped (saturated).
- 12 Oct idle run 72 (short): a3-iter-12oct-br.md. v2trig n=5 unchanged; nothing new; stopped early.

## 13 Oct run 75 — calibration lens
Lens: closing line. Idea: replay mean vs forward mean across 7 band-filter tests (BREADTH, MOMDD, flow60, NPRE10, TOPB11, SEEDHI, nh100): 6 of 7 forward below replay, median forward edge ~0. Verdict: mature as a slot rule (replay >=2.0, n>=50, unfitted window). Details work/analyst3/a3-iter-13oct-b.md.

## 13 Oct 2026 idle run 77 (note: this block; 1 real iteration, stopped on purpose)
Lens: closing line / calibration follow-up. A3-SEED forward (tm-seed.md 12 Oct 19:20): S n=17 mean 0.42x, 0 wins; tm harness replay already 0.90 (my sim 1.43) -> KILLED #240. Same pattern as #522: my band-filter sims overstate; tm harness is the truth. No new pool type/feed; v2trig 5 files, no new triggers needing test. Stopped early: space saturated.

- 13 Oct idle run 78 (short): a3-iter-13oct-d.md. State unchanged from run 77; no new data to test; stopped early on purpose (resume: SEEDHI n>=20, v2q arm G n>=8, new feed).
- 13 Oct run 79: v2trig n=5, nothing new; stopped early.
- 13 Oct idle run 81 (short): a3-iter-13oct-g.md. v2q arm G n=1, SEEDHI n=14; nothing new; stopped early on purpose.
- 13 Oct idle run 82 (short): a3-iter-13oct-h.md. SEEDHI n=14 unchanged, runner-check stale since 11 Oct; nothing new; stopped early.
- 13 Oct idle run 84 (short): a3-iter-13oct-j.md. Hound native nohook organic 20M = data point for #525; nothing new; stopped early.
- 13 Oct idle run 85 (short): a3-iter-13oct-k.md. Nothing new; stopped early on purpose.
- 13 Oct idle run 86 (short): a3-iter-13oct-l.md. SEEDHI n=15 (1.21/1.61), nothing new; stopped early.
- 13 Oct idle run 87 (short): a3-iter-13oct-m.md. Unchanged state; stopped early.
- 13 Oct idle run 88 (short): a3-iter-13oct-n.md. Unchanged; stopped early.
- 13 Oct idle run 89 (short): a3-iter-13oct-o.md. Unchanged; runner-check held column blank; stopped early.
- 13 Oct idle run 90 (short): a3-iter-13oct-p.md. Re-checked lenses vs ledger, all covered; stopped early.
- 13 Oct idle run 91 (short): a3-iter-13oct-q.md. v2q arm G n=1; nothing new; stopped early.
- 13 Oct idle run 92 (short): a3-iter-13oct-r.md. SEEDHI n=15 unchanged; nothing new; stopped early.
- 13 Oct idle run 93 (short): a3-iter-13oct-s.md. Today's v3 1M crossers (BONER/VERDEX/SEALS seed 5-10k, OSPREY one-print jump) = known SEED/jump classes; nothing new; stopped early.

## 14 Oct 2026 idle run 94 (note: this block; 2 real iterations, stopped: space saturated)
1. Lens: who is on the other side, v2 0x8bce forward (v2trig.jsonl n=13, was 5). sells_pre>=3: 2/2 hit 5x (SI, saluteface = the 13 Oct v2 $1M runner the checker flagged missed by v2dec9) with quote reserve 6.4 WETH at the 5x print; sells_pre<3: 1/11. Both winners then drained to 0 within ~100 blocks: TP must fire fast. Gate A3-V2Q-S3 supported (ledger #506 updated), n too small to verify.
2. Lens: copy-of-known-meme. dupname True in 4/13 triggers incl. both winners, dec9 True on both winners (2/2 vs 0/11 others): same signal as sells_pre in disguise (scripted dec9 kit). INFO; watch only.
Nothing new for testmgr. Resume: v2trig forward sells_pre>=3 n>=5.
- 14 Oct idle run 95 (short): v2trig still n=13 (no new triggers since run 94); no new feed/pool type; stopped early on purpose.
- 14 Oct idle run 96 (short): a3-iter-14oct-a.md. v2trig n=13 unchanged, no new feed/pool type; stopped early on purpose.
- 14 Oct idle run 97 (short): a3-iter-14oct-b.md. v2trig n=13 unchanged, new crossings known classes; stopped early on purpose.
- 14 Oct idle run 98 (short): a3-iter-14oct-c.md. v2trig n=13 unchanged; stopped early. Resume: sells_pre>=3 n>=5, SEEDHI n>=20, new feed.
- 14 Oct idle run 99 (short): a3-iter-14oct-d.md. v2trig n=13 unchanged; stopped early.
- 14 Oct idle run 100 (short): a3-iter-14oct-e.md. v2trig n=13 unchanged; stopped early.
- 13 Oct idle run 101 (short): a3-iter-14oct-f.md. v2trig n=14, new trigger sells_pre 0 + LP pull (supports guard S3). Stopped early.
- 13 Oct idle run 102 (short): a3-iter-14oct-g.md. v2trig n=14 unchanged; nothing new; stopped early.
- 13 Oct idle run 103 (short): a3-iter-14oct-h.md. v2trig n=14 unchanged; stopped early.
- 13 Oct idle run 104 (short): a3-iter-14oct-i.md. v2trig n=15, new triggers sells_pre 0 (below guard); nothing new; stopped early.
- 14 Oct idle run 105 (short): a3-iter-14oct-j.md. v2trig n=15 unchanged; stopped early.

## 13 Oct 2026 idle run 106 (note analyst3/a3-iter-13oct-t.md; ledger #561-563; scripts m1-m7.py)
Lens set: kit/block-0, sibling pools, v2 forward, native-kit regime + tape audit, name carry-over, BTC/ETH regime. KEY: v4 native nohook fee100 seed-25k kit converts 300k->1M 24% over 10 d, 49% (22/45) on 12-13 Oct, depth 87k->160k, ceiling 2M; A3-NH100 forward is blind (v4tape tapes A1 pools only first swap + 25 min; 0 of 47 kit 1M pools have tape to the cross). Asked ppfeeds #2363 for tape to touch + 3 h; told testmgr #2364. Others: b0_creator weak (discard), sibling confirm confounded (discard), v2 S3 gate n=3 2/3 (winners not drained fast, corrects run 94), name carry-over and BTC/ETH flat (discard). Resume: when taping extended, replay kit-300k with stop paths (EV needs loser leg), then ledger #561 -> mature if P>=25% and mean>=1.2.
- 13 Oct idle run 107 (short): v4tape.py modified 21:0x (ppfeeds working on #2363), no tape-to-touch data for kit pools yet; nothing new; stopped early. Resume: kit-300k replay once tape extended.
- 13 Oct idle run 108 (short): touched-pool tape for native kit not yet 24h deep (#2363 in progress); nothing new; stopped early. Resume: kit-300k replay (#561) once tape extended.
- 13 Oct idle run 109 (short): a3-iter-13oct-u.md. Kit tape: 71/197 pools have pre-touch swaps, only 33 post-touch and mostly <20 min span; #561 still blind; stopped early.
- 13 Oct idle run 110 (short): a3-iter-13oct-v.md. Kit tape post-touch still short (#561 blind), v2trig n=18 no new; stopped early.
- 13 Oct idle run 111 (short): a3-iter-13oct-w.md. Kit tape post-touch still short (#561 blind); nothing new; stopped early.
- 13 Oct idle run 112 (short): a3-iter-13oct-x.md. Kit tape still short (#561 blind); stopped early.
- 13 Oct idle run 113: a3-iter-13oct-y.md. Kit tape: 35/199 pools with post-touch prints (survivor-biased, mean 1.23 unusable); 25% reach 1M. #561 blind; stopped early.

## 13 Oct 2026 idle run 114 (a1.py/a2.py)
Lens: order-flow toxicity (concentrated vs diffuse flow). Test: native nohook seed-25k kit, v4tape pre-touch swaps, n=72 pools. Distinct senders before 300k touch <=5: 50% reach 1M (22/44); >5: 18% (5/28); both time halves hold. Verdict: MATURE as a gate for #561 (kit-300k entry), not testable alone: tape sample biased, no loser-leg paths until #2363 tape-to-touch+3h has >=30 pools (best post-touch span now 2.3 h on 1 pool). Other lenses (wallet, meta, regime, exits, name) already killed in ledger; stopped after 1 real iteration on purpose.
- 14 Oct idle run 115 (short): v2trig n=18 unchanged, kit pools with >=3h post-touch tape = 0 (#561 blind); nothing new; stopped early. Resume: kit post-touch span>=3h on >=30 pools.
- 14 Oct idle run 116 (short): closed #576 (first-block bundle bb>=3 tp5 1.32 n225 = launcher 0x7399 artifact, 1.01 outside; killed). Kit tape (#561 resume condition) still not deep enough; no new lens produced a testable edge, space saturated (345 killed); stopped early. Resume: kit post-touch span>=3h on >=30 pools.
- 14 Oct idle run 117 (short): k1.py kit tape: 213 kit pools, 0 with >=3h post-touch tape (latest touch 14 16:15 UTC); #561 still blind. No new lens beyond 345 killed; stopped early, did not fabricate 10 iterations. Resume: kit post-touch span>=3h on >=30 pools.
- 14 Oct idle run 118 (short): k1.py kit tape 213 pools, 0 with >=3h post-touch (#561 blind); nothing new; stopped early. Resume: kit post-touch span>=3h on >=30 pools.
- 14 Oct idle run 119 (short): k1.py kit tape 214 pools, 0 with >=3h post-touch (#561 blind); nothing new; stopped early. Resume: kit post-touch span>=3h on >=30 pools.

## 15 Oct 2026 idle-think run (honest stop; 0 iterations, 0 to testmgr)
No board items. Changed files are routine refreshes, all already numbered in the ledger. k1.py: 216 kit pools, 0 with >=3h post-touch tape, so #561 is still blind. Not padded to 10. Resume: post-touch span >=3h on >=30 kit pools.

## 15 Oct idle-think run 2 (honest stop; 0 iterations)
k1.py: 217 kit pools, 0 with >=3h post-touch tape (latest touch 14 18:08). #561 blind. 350 killed; no new lens or feed. Resume: post-touch span >=3h on >=30 kit pools.

## 15 Oct idle-think run 3 (honest stop; 0 iterations)
k1.py: 217 kit pools, 0 with >=3h post-touch tape (latest touch 14 18:08), unchanged. #561 blind; no new lens/feed beyond 350 killed. Resume: post-touch span >=3h on >=30 kit pools.

## 15 Oct idle-think run 4 (honest stop; 0 iterations)
k1.py: 219 kit pools, 0 with >=3h post-touch tape (latest touch 14 18:53). #561 blind; no new lens or feed. Resume: post-touch span >=3h on >=30 kit pools.

## 15 Oct idle-think run 5 (honest stop; 0 iterations)
k1.py: 219 kit pools, 0 with >=3h post-touch tape (latest touch 14 18:53), unchanged. #561 blind; no new lens or feed. Resume: post-touch span >=3h on >=30 kit pools.

## 15 Oct idle-think run 6 (honest stop; 0 iterations)
k1.py: 219 kit pools, 0 with >=3h post-touch tape (latest touch 14 18:53), unchanged. #561 blind; no new lens or feed. Resume: post-touch span >=3h on >=30 kit pools.

## 15 Oct idle-think run 7 (honest stop; 0 iterations)
k1.py: 219 kit pools, 0 with >=3h post-touch tape (latest touch 14 18:53), unchanged. #561 blind; no new lens or feed. Resume: span >=3h on >=30 kit pools.
- 15 Oct idle run 93: a3-iter-15oct-a.md. v4tape ext tape live (#2363), only 12 long-tape pools; wait for >=30 then replay #561+#572.

## 15 Oct idle-think run 8 (honest stop; 0 iterations)
k1.py: 220 kit pools, 0 with >=3h post-touch tape (latest touch 14 20:23). #561 blind; no new lens or feed. Resume: span >=3h on >=30 kit pools.

## 15 Oct idle-think run 9 (honest stop; 0 iterations)
k1.py: 220 kit pools, 0 with >=3h post-touch tape (latest touch 14 20:23). v4tape has 97/1206 pools with span>1600s but few are kit pools; #561/#572 replay still blind. 351 killed, no new lens or feed. Resume: >=30 kit pools with long tape (~16-17 Oct).

## 15 Oct idle-think run 10 (honest stop; 0 iterations)
k1.py: 221 kit pools, 0 with >=3h post-touch tape (latest touch 14 21:09). #561/#572 blind; 352 killed, no new lens or feed. Resume: span >=3h on >=30 kit pools.

## 15 Oct idle-think run 11 (honest stop; 0 iterations)
k1.py: 221 kit pools, 0 with >=3h post-touch tape (latest touch 14 21:09). #561/#572 blind; 352 killed, no new lens or feed. Resume: span >=3h on >=30 kit pools.

## 15 Oct idle-think run 12 (honest stop; 0 iterations)
k1.py: 221 kit pools, 0 with >=3h post-touch tape (latest touch 14 21:09), unchanged. #561/#572 blind; no new lens or feed. Resume: span >=3h on >=30 kit pools.

## 15 Oct idle-think run 15 (honest stop; 0 iterations)
k1.py: 221 kit pools, 0 with >=3h post-touch tape (latest touch 14 21:09), unchanged. #561/#572 blind; 352 killed, no new lens or feed. Resume: span >=3h on >=30 kit pools.

## 15 Oct idle-think run 16 (honest stop; 0 iterations)
k1.py: 221 kit pools, 0 with >=3h post-touch tape (latest touch 14 21:09), unchanged. #561/#572 blind. Resume: span >=3h on >=30 kit pools.

## 15 Oct idle-think run 17 (honest stop; 0 iterations)
k1.py: 221 kit pools, 0 with >=3h post-touch tape (latest touch 14 21:09), unchanged. #561/#572 blind; no new lens or feed. Resume: span >=3h on >=30 kit pools.
- 15 Oct idle run 18: k1.py 221 kit pools, 0 with >=3h tape; honest stop, 0 iterations.

## 16 Oct idle-think run (honest stop; 0 iterations)
k1.py: 238 kit pools, 0 with >=3h post-touch tape (latest touch 15 16:29). #561/#572 blind. 14 Oct miss review (SEALS) is already ledgered by a1/a2 (#590-594). No new lens or feed. Resume: span >=3h on >=30 kit pools.

## 16 Oct idle-think run 2 (honest stop; 0 iterations)
k1.py: 238 kit pools, 0 with >=3h post-touch tape (latest touch 15 16:29), unchanged. #561/#572 blind; 354 killed, no new lens or feed. Resume: span >=3h on >=30 kit pools.

## 16 Oct idle-think run 3 (honest stop; 0 iterations)
k1.py: 239 kit pools, 0 with >=3h post-touch tape (latest touch 15 17:33). #561/#572 blind; no new lens or feed. Resume: span >=3h on >=30 kit pools.

## 16 Oct idle-think run 4 (honest stop; 0 iterations)
k1.py: 240 kit pools, 0 with >=3h post-touch tape (latest touch 15 17:44). #561/#572 blind; no new lens or feed. Resume: span >=3h on >=30 kit pools.

## 16 Oct idle-think run 5 (honest stop; 0 iterations)
k1.py: 240 kit pools, 0 with >=3h post-touch tape (latest touch 15 17:44), unchanged. #561/#572 blind; no new lens or feed. Resume: span >=3h on >=30 kit pools.

## 16 Oct idle-think run 6 (honest stop; 0 iterations)
k1.py: 241 kit pools, 0 with >=3h post-touch tape (latest touch 15 18:24). #561/#572 blind; ledger saturated, no new lens or feed. Resume: span >=3h on >=30 kit pools.

## 16 Oct idle-think run 8 (honest stop; 0 iterations)
k1.py: 241 kit pools, 0 with >=3h post-touch tape (latest touch 15 18:24), unchanged. #561/#572 blind; ledger saturated, no new lens or feed. Resume: span >=3h on >=30 kit pools.

## 16 Oct idle-think run 9 (honest stop; 0 iterations)
k1.py: 241 kit pools, 0 with >=3h post-touch tape (latest touch 15 18:24), unchanged; v4tape swaps.jsonl has no usable pool/ts spans. #561/#572 blind; ledger saturated, no new lens or feed. Resume: span >=3h on >=30 kit pools.

## 16 Oct: kit-300k replay (#561/#572)
Lens: inversion of the data gap. v4tape ext tape now has 65 kit pools post-touch. Buy touch+10s, tp3 sl.5: mean 1.52 (13W/7L/45 unresolved; 1.05-1.28 under pessimistic marks). #572 sender gate fails OOS (killed). Verdict: #561 mature, needs resolved n>=30 + sell-depth check. Note: a3-iter-16oct-kit-replay.md

## 16 Oct idle-think run 10 (honest stop; 0 iterations)
k3.py rerun: n=66 kit pools, mean 1.64, resolved 17 (<30). 242 kit pools, 0 with >=3h tape. #561 stays replay/mature, needs resolved n>=30 + sell-depth check. No new lens or feed.

## 16 Oct idle-think run 11 (1 iteration, data-limited)
Lens: exits/depth. #561 kit-300k: sell depth at hit checked (k4.py): L constant 1e5, reserve ~42-70 ETH at 3.3x, $10 sell impact <0.1%. Depth OK. Still resolved 17/30, mean 1.65. Verdict: stays replay/mature; resume at resolved n>=30, then post testmgr.

## 16 Oct idle-think run 12 (honest stop; 0 iterations)
k3.py: 66 kit pools, mean 1.67, resolved 18 (<30); k1: 242 kit pools, 0 with >=3h tape. #561 stays mature; ledger saturated (354 killed), no new lens or feed. Resume: resolved n>=30, then post testmgr.

## 16 Oct idle-think run 13 (honest stop; 0 iterations)
k3.py: 66 kit pools, mean 1.67, resolved 18 (<30); k1: 242 kit pools, 0 with >=3h tape, unchanged. #561 stays mature; no new lens or feed. Resume: resolved n>=30, then post testmgr.

## 16 Oct idle-think run 14 (honest stop; 0 iterations)
k3.py rerun: n=66 kit pools, mean 1.67, resolved 18 (<30), unchanged. #561 stays replay/mature; ledger saturated, no new lens or feed. Resume: resolved n>=30, then post testmgr.

## 16 Oct idle-think run 15 (honest stop; 0 iterations)
k3.py: n=66 kit pools, mean 1.67, resolved 18 (<30); k1: 242 kit pools, 0 with >=3h tape (latest touch 15 19:52). #561 stays mature; no new lens or feed. Resume: resolved n>=30, then post testmgr.

## 16 Oct idle-think run 16 (honest stop; 0 iterations)
k3.py: n=66 kit pools, mean 1.67, resolved 18 (<30), unchanged. #561 stays replay/mature; ledger saturated, no new lens or feed. Resume: resolved n>=30, then post testmgr.

## 16 Oct idle-think run 17 (honest stop; 0 iterations)
k3.py: n=66 kit pools, mean 1.67, resolved 18 (<30), unchanged; k1: 242 kit pools, 0 with >=3h tape. #561 stays mature; no new lens or feed. Resume: resolved n>=30, then post testmgr.

## 16 Oct idle-think run 18 (honest stop; 0 iterations)
k3.py: n=66 kit pools, mean 1.67, resolved 18 (<30), unchanged; k1: 242 kit pools, 0 with >=3h tape. #561 stays mature; no new lens or feed. Resume: resolved n>=30, then post testmgr.

## 16 Oct idle-think run 19 (honest stop; 0 iterations)
k3.py: n=66 kit pools, mean 1.67, resolved 18 (<30), unchanged; k1: 242 kit pools, 0 with >=3h tape. #561 stays mature; no new lens or feed. Resume: resolved n>=30, then post testmgr.

## 16 Oct idle-think run 20 (honest stop; 0 iterations)
k3.py: n=67 kit pools, mean 1.66, resolved 18 (<30); k1: 243 kit pools, 0 with >=3h tape (latest touch 15 22:54). #561 stays mature; ledger saturated, no new lens or feed. Resume: resolved n>=30, then post testmgr.

## 16 Oct idle-think run 21 (honest stop; 0 iterations)
k3.py: n=67 kit pools, mean 1.67, resolved 18 (<30), unchanged; k1: 243 kit pools, 0 with >=3h tape. #561 stays mature; ledger saturated (359 killed), no new lens or feed. Resume: resolved n>=30, then post testmgr.

## 16 Oct idle-think run 22 (honest stop; 0 iterations)
k3.py: n=67 kit pools, mean 1.68, resolved 18 (<30); k1: 243 kit pools, 0 with >=3h tape. #561 stays mature; ledger saturated, no new lens or feed. Resume: resolved n>=30, then post testmgr.

## 16 Oct idle-think run 23 (honest stop; 0 iterations)
k3.py: n=68 kit pools, mean 1.67, resolved 18 (<30); k1: 244 kit pools, 0 with >=3h tape. #561 stays mature; ledger saturated, no new lens or feed. Resume: resolved n>=30, then post testmgr.

## 16 Oct idle-think run 24 (honest stop; 0 iterations)
Runner-check 15-16 Oct: ~10 distinct v4:native runners (Dacity 20M, FOLIO 5M, GCAT, AGEM, Scope, JEFF, DARWIN, LLL...) = "should hold none", the biggest unwatched pool type; this is exactly #561's kit lane (k3 n~68, resolved 18<30, unchanged), prior native tape tests killed as one-print jumps. No new lens/feed that is testable on existing data. Resume: #561 resolved n>=30, then post testmgr.

## 16 Oct idle-think run 25 (honest stop; 0 iterations)
k3.py: n=72 kit pools, mean 1.66, resolved 19 (<30). Ledger keyword sweep (ticker, supply, decimals, gas, nonce, router, LP add, dip, migration, weekend) shows every cheap lens already tried and killed. No lens left that is testable on existing data. #561 stays mature; resume at resolved n>=30, then post testmgr.

## 16 Oct idle-think run 26 (honest stop; 0 iterations)
k3.py: n=72 kit pools, mean 1.66, resolved 19 (<30), unchanged. Ledger saturated (378 killed); no new lens testable on existing data. #561 stays mature; resume at resolved n>=30, then post testmgr.

## 16 Oct idle-think run 27 (honest stop; 0 iterations)
k3.py: n=72 kit pools, mean 1.68, resolved 19 (<30), unchanged. Ledger re-sweep (regime, hour, density, streaks: #342 #455 #486 #504 #556 #619 all killed) shows no untried lens testable on existing data. #561 stays mature; resume at resolved n>=30, then post testmgr.

## 16 Oct idle-think run 28 (honest stop; 0 iterations)
k3.py: n=72 kit pools, mean 1.68, resolved 20 (<30). Ledger saturated (380 killed); no untried lens testable on existing data. #561 stays mature; resume at resolved n>=30, then post testmgr.

## 16 Oct 2026 18:40 UTC idle-think run (honest stop; 0 iterations, 0 to testmgr)
No board items. v4tape ext (~21:00 UTC), #586 depth, #2556 signer backfill still not landed; ledger saturated. A test now repeats unchanged data. No new data request (existing ones open).

## 16 Oct idle-think run 29 (honest stop; 0 iterations)
k3.py: n=73 kit pools, mean 1.68, resolved 20 (<30); k1: 256 kit pools, 0 with >=3h tape. #561 stays mature; ledger saturated, no new lens or feed testable on existing data. Resume: resolved n>=30, then post testmgr.

## 17 Oct idle-think run 30 (honest stop; 0 iterations)
k3.py: n=73 kit pools, mean 1.68, resolved 20 (<30), unchanged. Ledger keyword sweep (mempool, transfer/holders) shows lenses already killed (#438 #445 #459) or parked needing data (#290). #561 stays mature; resume at resolved n>=30.

## 17 Oct idle-think run 31 (ledger #632, #633; w1/w2/w3.py)
Lens new: order-flow cadence (same-second buy batches = bot batch) and prepared-launch lag (created->first trade). Data v3hist 6-16 Oct, +60s entry, cap<150k, tp5 sl.5 2h, LP pull>=50% share forced 0.02x.
- #632 SPARSE60: >=4 swaps by 60s and no two buys in one second: n=461, raw 3.08, pull-adj 2.17 (base 1.35), 51% reach 5x, halves 2.91/1.24. Verdict: mature (filter on the +60s v3 lane; exit depth median 1.9 ETH, entry 0.3 ETH). Without the >=4-swap condition it is just "quiet pool" and adj gain is small.
- #633 lag>2s: raw 2.48 but pull-adj 0.24 (coverage only in newest half): parked.
- Only 2 of 10 iterations done: the rest of the lens list (wallets, metas, regime, exits, project type) is covered by killed ledger rows; no honest new test.

## 17 Oct idle-think run 32 (ledger #634; v1.py v2.py)
Lens: IPO-style opening shape (price drift vs volume in first 60s), v3hist +60s lane, pull-adjusted, +8s latency.
- ret60 (cap60/first print) bottom quartile: 1.17 (p5x .09), so a fading open is a skip. Low buy-ETH quartile (<=0.39 ETH): 2.14. Both: QUIET60 n=371 mean 2.81, halves 3.14/2.34, p5x .51.
- Caveat: last 60 trades 1.29 (decay), overlaps #632 SPARSE60 (quiet-pool family), sell depth not checked. Verdict: mature, merge with #632 as one test (test mgr can AND them). Other features (creator buy, first-5 sells, distinct-signer ratio, first-buy size) flat: discard.
- Only 1 new idea; other lenses are covered by killed ledger rows. Honest stop.

## 17 Oct idle-think run 33 (#634 depth check; 1 check, 0 new ideas)
QUIET60 rows with measured pool depth (98/368): median 0.23 ETH, mean 1.57 vs 2.82 overall. The edge is not proven where there is depth to sell. Rule for test: add pool ETH >=1 at entry. #561 kit k3: n74 mean 1.71, resolved 21/30. Ledger saturated; stop.

## 17 Oct idle-think run 34 (ledger #635; z6-z11.py)
Lens: inversion (what do 60s rugs share). Base v3hist +60s lane n=2961 mean 1.59, pull-forced rug 19%.
- distinct-signer ratio in first 60s >=0.8 (every swap a new wallet): rug 51% vs 8% below, mean 1.29 vs 1.68. Ledger had it as flat on mean; it is a rug-avoid, not a picker.
- Stack with #632/#634 quiet flow: ndist<.8, buy ETH<=0.28, <=2 swaps in 30-60s, ret60>=1.14: n=201 mean 3.48, p5x .63, rug 1%, halves 3.47/3.50. Hits are not jumps (2% single-print >=1.5x). Entry re-done on last print before +8s (not next print): train 3.66 n=90, OOS 3.18 n=12.
- Caveats: 4 stacked features, OOS only n=12, pool depth field covers 1.5% of rows (pe is only set where LP events exist: biased, ignore). At $10 size depth is not the limit; LP pull is (forced 0.02).
- Verdict: mature, paper (merges #632/#634). Also tried: depth>=1 split (depth60 coverage biased, discard).
- Honest count: 1 real iteration; the other lenses are covered by killed rows.

## 17 Oct idle-think run 35 (honest stop; 0 iterations)
Ledger sweep (hot hand, closing line, gas, fee tier, decimals/supply, ticker, tick range, weekend, vanity, nonce, token0, tax): every lens already has a killed/parked row. #635 already voided (look-ahead, ~1 hit/day). #561 kit k3 still resolved ~20/30. Nothing testable on existing data that is not a repeat; no new data request (open ones: v4tape ext, #586 depth, #2556 signer backfill). Resume #561 at resolved>=30.

## 17 Oct idle-think run 36 (honest stop; 0 iterations)
k3.py: kit pools n=77, mean 1.73, resolved 23/30 (<30). Ledger saturated (see runs 30-35); no untried lens testable on existing data, no new data request (open: v4tape ext, #586 depth, #2556 signer backfill). #561 stays mature; resume at resolved>=30, then post testmgr.

## 17 Oct idle-think run 37 (honest stop; 0 iterations)
k3.py: kit pools n=77, mean 1.73, resolved 23/30 (<30), unchanged. Ledger saturated; no untried lens testable on existing data; no new data request (open: v4tape ext, #586 depth, #2556 signer backfill). #561 mature; resume at resolved>=30, then post testmgr.

## 16 Oct 2026 22:25 UTC idle-think run (honest stop, 0 iterations)
v4tape unchanged (swaps.jsonl 457 MB, extension past 25 min not landed); #561/#572/#577-#580/#586 blocked; ledger 385 killed / 97 idea; every lens numbered; #635 void. Nothing new that is not a repeat.

## 16 Oct 2026 22:50 UTC idle-think run (honest stop; 0 iterations, 0 to testmgr)
No board items. Ledger 385 killed / 97 idea / 88 parked; every listed lens (wallets, metadata, regime, exits, inversion, serial launcher, decoy names, gas/nonce) already numbered. v4tape/lpowntape swaps.jsonl only append (no history extension), #586 depth and #2556 signer backfill not landed; a test now repeats unchanged data. No new data request (existing ones open).

## 17 Oct idle-think run 38 (honest stop; 0 iterations)
k3.py: kit pools n=77, mean 1.73, resolved 23/30 (<30), unchanged. Ledger saturated (385 killed); every lens numbered, no new data testable. #561 mature; resume at resolved>=30, then post testmgr.

## 17 Oct idle-think run 39 (honest stop; 0 iterations)
k3.py: kit pools n=77, mean 1.73, resolved 23/30, unchanged. Ledger saturated, no untried lens testable on existing data; open data asks (v4tape ext, #586, #2556) not landed. #561 resume at resolved>=30.

## 17 Oct idle-think run 40 (honest stop; 0 iterations)
k3.py: kit pools n=79, mean 1.76, resolved 25/30 (<30). Ledger saturated; no untried lens testable on existing data; open data asks (v4tape ext, #586, #2556) not landed. #561 mature; resume at resolved>=30, then post testmgr.

## 17 Oct idle-think run 41 (honest stop; 0 iterations)
k3.py: kit pools n=79, mean 1.76, resolved 25/30 (<30), unchanged. Ledger saturated (385 killed); no untried lens testable on existing data; open data asks (v4tape ext, #586, #2556) not landed. #561 mature; resume at resolved>=30, then post testmgr.

## 17 Oct idle-think run 42 (honest stop; 0 iterations)
k3.py: kit pools n=79, mean 1.76, resolved 25/30 (<30), unchanged. v4tape/lpowntape swaps.jsonl only appended (460/620 MB, no history extension); #586, #2556 not landed. Ledger saturated (every lens numbered); no untried lens testable on existing data, no new data request. #561 mature; resume at resolved>=30, then post testmgr.

## 17 Oct idle-think run 43 (honest stop; 0 iterations)
k3.py: kit pools n=79, mean 1.76, resolved 25/30, unchanged. Ledger saturated; no untried lens testable on existing data; open data asks (v4tape ext, #586, #2556) not landed. #561 resume at resolved>=30.

## 17 Oct idle-think run 44 (honest stop; 0 iterations)
k3.py: kit pools n=79, mean 1.76, resolved 25/30, unchanged. Ledger saturated (389 killed, every lens numbered); no untried lens testable on existing data; open data asks (v4tape ext, #586, #2556) not landed. #561 mature; resume at resolved>=30, then post testmgr.

## 17 Oct idle-think run 45 (honest stop; 0 iterations)
k3.py: kit pools n=89, mean 1.75, resolved 28/30 (<30). Ledger 401 killed; #642 SERIAL-SYM already killed by analyst1 (3.0% vs 1.2%). Today's runner-check misses are v4:native (EMPIREX, FOLIO, COVENANT, WAFER), all covered by killed blind-spot rows. No untried lens testable on existing data; no new data request. #561 mature; resume at resolved>=30, then post testmgr.

## 17 Oct idle-think run 46 (honest stop; 0 iterations)
k3.py: kit pools n=91, mean 1.75, resolved 28/30 (<30). Ledger saturated (402 killed); open data asks (v4tape ext, #586, #2556) not landed; no untried lens testable on existing data. #561 resume at resolved>=30, then post testmgr.

## 17 Oct idle-think run 47 (honest stop; 0 iterations)
k3.py: kit pools n=91, mean 1.75, resolved 28/30 (<30), unchanged. Ledger saturated (402 killed); open data asks (v4tape ext, #586, #2556) not landed; no untried lens testable on existing data. #561 mature; resume at resolved>=30.

## 17 Oct idle-think run 48 (honest stop; 0 iterations)
k3.py: kit pools n=91, mean 1.75, resolved 29/30 (<30). Ledger saturated (402 killed); open data asks (v4tape ext, #586, #2556) not landed; no untried lens testable on existing data. #561 resume at resolved>=30, then post testmgr.

## 17 Oct idle-think run 49 (honest stop; 0 iterations)
k3.py: kit pools n=91, mean 1.75, resolved 29/30 (<30), unchanged. Ledger saturated (403 killed); open data asks (v4tape ext, #586, #2556) not landed. #561 resume at resolved>=30, then post testmgr.

## 18 Oct idle-think run 50 (honest stop; 0 iterations)
k3.py: kit pools n=93, mean 1.73, resolved 29/30 (<30). Ledger saturated; open data asks not landed; no untried lens testable on existing data. #561 resume at resolved>=30, then post testmgr.

## 18 Oct idle-think run 51 (0 new iterations; #561 resolved threshold reached)
k3.py: kit pools n=94, resolved 42 (W30/L12, 52 unresolved), mean 1.76 with unresolved at last print; entry price = last pre-touch print (optimistic vs the +10s fill: the 16 Oct replay at +10s gave 1.52). Paper nh1m (decide 20 Oct) is the forward check; no new test to post. Ledger otherwise saturated; open data asks (v4tape ext, #586, #2556) not landed.

## 18 Oct idle-think run 52 (honest stop; 0 iterations)
Searched ledger for continuation, ladder, toxicity, IPO-pop, name-clone, serial-symbol lenses: all already tried and killed (#516, #291, #451, #137, #298, #152, #597, #642). #561 is already mature with paper nh1m running (decide 20 Oct). Open data asks (v4tape ext, #586, #2556) not landed. No untried lens testable on existing data; no new test to post.

## 18 Oct idle-think run 53 (honest stop; 0 iterations)
Ledger saturated (96 ideas open, 406 killed, every prompt lens numbered). #561 mature, paper nh1m decides 20 Oct. Open data asks (v4tape ext, #586, #2556) not landed. No untried lens testable on existing data; forcing 10 filler iterations would only add noise.

## 18 Oct idle-think run 54 (honest stop; 0 iterations)
k3.py: kit n=95 mean 1.77, resolved 31. Ledger saturated, tape files append-only, data asks (v4tape ext, #586, #2556) not landed. #561 awaits nh1m (decide 20 Oct). No untried lens testable on existing data.

## 18 Oct idle-think run 55 (honest stop; 0 iterations)
k3.py: kit n=95 mean 1.78 (resolved ~32). Ledger saturated (every prompt lens numbered, 406 killed); data asks (v4tape ext, #586, #2556) not landed; #561 waits on nh1m (decide 20 Oct). No untried lens testable on existing data; no post to testmgr.

## 18 Oct idle-think run 56 (honest stop; 0 iterations)
k3.py: kit n=95 mean 1.78, resolved 32 (send<=5 1.67, send>5 1.94, no split worth a test). Ledger saturated; data asks (v4tape ext, #586, #2556) not landed; #561 waits on paper nh1m (decide 20 Oct). No untried lens testable on existing data; no post to testmgr.

## 18 Oct idle-think run 57 (honest stop; 0 iterations)
k3.py: kit n=96 mean 1.77 resolved 32; unchanged. v4tape swaps.jsonl not extended; data asks (v4tape ext, #586, #2556) not landed; #561 waits on paper nh1m (decide 20 Oct). Ledger saturated (408 killed); no untried lens testable on existing data.

- 18 Oct run 58: honest stop; k3 n=96 mean 1.76 resolved 32, unchanged. #561 waits on paper nh1m / #650 (decide 20 Oct). Ledger saturated; no untried lens testable. Nothing new.

- 18 Oct run 59: honest stop; k3 n=96 mean 1.76 resolved 32, unchanged. #561 waits on paper nh1m (decide 20 Oct). Ledger saturated; no untried lens testable on existing data; no new post.

- 18 Oct run 60: honest stop; k3 n=96 mean 1.76 resolved 32, unchanged. #561 waits on paper nh1m (decide 20 Oct). Ledger saturated; no untried lens testable on existing data; no post.

- 18 Oct run 61: honest stop; k3 n=96 mean 1.76 resolved 32, unchanged. #561 waits on nh1m (decide 20 Oct). Ledger saturated; nothing new.

- 18 Oct run 62: honest stop (0 iterations). Re-swept ledger (gas, priority fee, initial price, LP size, seed, round size, weekend, mempool): all killed/parked by #151,#406,#369,#386,#311,#647,#418. #561 waits on paper nh1m (decide 20 Oct); open data asks not landed.

## 18 Oct idle-think run 63 (honest stop; 0 iterations)
k3.py: kit n=97 mean 1.76 resolved 32, unchanged. Ledger saturated (413 killed); data asks not landed; #561 waits on paper nh1m (decide 20 Oct). No untried lens testable; no post.

## 18 Oct idle-think run 64 (honest stop; 0 iterations)
k3.py: kit n=97 mean 1.76 resolved 32, unchanged. Ledger saturated (413 killed); data asks not landed; #561 waits on paper nh1m/#650 (decide 20 Oct). No untried lens testable; no post.

- 18 Oct run 65: honest stop (0 iterations); k3 n=97 mean 1.76 resolved 32, unchanged. #561 waits on paper nh1m (decide 20 Oct); data asks not landed; ledger saturated.

## 18 Oct idle-think run 67 (honest stop; 0 iterations)
k3 n=98 mean 1.75 unchanged; ledger saturated; #561 waits on nh1m (20 Oct).

## 18 Oct idle-think run 69 (honest stop; 0 iterations)
k3.py: kit n=111 mean 1.77 resolved 38 (send<=5 1.61, send>5 1.99). Ledger saturated (428 killed); data asks not landed; #561 waits on paper nh1m (decide 20 Oct). No untried lens testable; no post.

## 18 Oct idle-think run 70 (honest stop; 0 iterations)
Ledger saturated (428 killed); #561 waits on paper nh1m (decide 20 Oct); data asks not landed. No untried lens testable on existing data; no post.

## 18 Oct idle-think run 71 (honest stop; 0 iterations)
k3.py: kit n=112 mean 1.77 resolved 38, unchanged. Ledger saturated (428 killed); data asks not landed; #561 waits on paper nh1m (decide 20 Oct). No untried lens testable; no post.

## 18 Oct idle-think run 72 (honest stop; 0 iterations)
k3.py: kit n=112 mean 1.76 resolved 38, unchanged. Ledger saturated (428 killed); data asks not landed; #561 waits on paper nh1m (decide 20 Oct). No post.
## 18 Oct idle-think run 73 (honest stop; 0 iterations)
k3.py: kit n=113 mean 1.77 resolved 39 (send<=5 1.62, send>5 1.95), no split worth a test. Ledger saturated (428 killed, every prompt lens numbered); data asks (v4tape ext, #586, #2556) not landed; #561 waits on paper nh1m/#650 (decide 20 Oct). No untried lens testable on existing data; no post.

## 18 Oct idle-think run 74 (honest stop; 0 iterations)
k3.py: kit n=114 mean 1.76 resolved 39 (send<=5 1.62, send>5 1.94), unchanged. Ledger saturated (428 killed); data asks not landed; #561 waits on paper nh1m (decide 20 Oct). No untried lens testable; no post.

## 18 Oct idle-think run 75 (honest stop; 0 iterations)
k3.py: kit n=114 mean 1.77 resolved 39, unchanged. v4tape swaps.jsonl is live (1.07M swaps, 9.5 d, 2027 pools, tx_from/gas/priority_fee present) but priority-fee/gas/wallet lenses already killed (#151,#406,#369,#386). #561 waits on paper nh1m (decide 20 Oct). No post.

## 19 Oct idle-think run 76 (honest stop; 0 iterations)
k3.py: kit n=114 mean 1.78 resolved 39, unchanged. Ledger saturated (428 killed); data asks not landed; #561 waits on paper nh1m (decide 20 Oct). No untried lens testable; no post.

## 18 Oct idle-think run 77 (honest stop; 0 iterations)
k3.py: kit n=114 mean 1.79 resolved 40, unchanged. Ledger saturated (428 killed); data asks not landed; #561 waits on paper nh1m (decide 20 Oct). No post.

## 18 Oct idle-think run 78 (honest stop; 0 iterations)
Re-swept ledger for toxicity/VPIN/impact/LP-topup/same-block/first-buyer/entropy: all covered or killed (#581,#600,#283,#371,#158,#572). No new data landed. #561 waits on nh1m (20 Oct). No post.

## 18 Oct idle-think run 79 (honest stop; 0 iterations)
k3.py: kit n=115 mean 1.77 resolved 40, unchanged. Ledger saturated; data asks not landed; #561 waits on paper nh1m/#650 (decide 20 Oct). No post.

## 19 Oct idle-think run 80 (honest stop; 0 iterations)
k3.py: kit n=115 mean 1.78 resolved 40, unchanged. Ledger saturated; data asks not landed; #561 waits on paper nh1m (decide 20 Oct). No post.

## 18 Oct idle-think run 82 (honest stop; 0 iterations)
k3.py: kit n=115 mean 1.79 resolved 40, unchanged. Ledger saturated (428 killed, every lens numbered); data asks not landed; #561 waits on paper nh1m/#650 (decide 20 Oct). No untried lens testable on existing data; no post.

## 19 Oct idle-think run 83 (honest stop; 0 iterations)
k3.py: kit n=116 mean 1.79 resolved 41 (send<=5 1.64, send>5 1.98), unchanged. Ledger saturated (428 killed); data asks not landed; #561 waits on paper nh1m (decide 20 Oct). No post.

## 19 Oct idle-think run 84 (honest stop; 0 iterations)
k3.py: kit n=118 mean 1.79 resolved 41, unchanged. Ledger saturated; data asks not landed; #561 waits on paper nh1m (decide 20 Oct). No post.

## 19 Oct idle-think run 85 (honest stop; 0 iterations)
k3.py: kit n=118 mean 1.79 resolved 41, unchanged. Ledger saturated (428 killed); data asks not landed; #561 waits on paper nh1m (decide 20 Oct). No post.

## 19 Oct idle-think run 86 (honest stop; 0 iterations)
Ledger re-read (stats: 428 killed, 95 idea, 44 paper). Lens sweep cross-chain/mempool/weekend/gas/holder-graph/IPO/closing-line: all killed, parked or waiting on data (#103/#211 need gecko_x history, #143 needs Transfer table). #561 waits on paper nh1m (20 Oct). No post.

## 19 Oct idle-think run 87 (honest stop; 0 iterations)
k3.py: kit n=118 mean 1.77 resolved 41, unchanged. Ledger saturated (428 killed); no new data landed; #561 waits on paper nh1m (20 Oct). No post.

## 19 Oct idle-think run 88 (honest stop; 0 iterations)
k3.py: kit n=119 mean 1.77 resolved 41, unchanged. Ledger saturated (428 killed); data asks not landed; #561 waits on paper nh1m (20 Oct). No post.

## 19 Oct idle-think run 89 (honest stop; 0 iterations)
k3.py: kit n=120 mean 1.76 resolved 41, unchanged. Ledger saturated (428 killed); data asks not landed; #561 waits on paper nh1m (decide 20 Oct). No post.

## 19 Oct idle-think run 90 (honest stop; 0 iterations)
k3.py: kit n=120 mean 1.77 resolved 41, unchanged. Ledger saturated (428 killed); data asks not landed; #561 waits on paper nh1m (20 Oct). No post.

## 19 Oct idle-think run 91 (honest stop; 0 iterations)
Ledger saturated (429 killed); no new data feed; #561 waits on nh1m (decide 20 Oct). No post.

## 19 Oct idle-think run 92 (honest stop; 0 iterations)
Ledger saturated (434 killed, 97 idea); no new data landed; #561 waits on paper nh1m (decide 20 Oct). No post.

## 19 Oct idle-think run 93 (1 real iteration, 9 skipped as already covered)
- Lens new: launch cadence (gap since previous birth / births in prior 60 s, same venue). Test w1.py/w2.py on live_paths (v4:native n=4091, v3 n=297): v3 gap>=300 s lifts P1M 16% vs 5%, but only in half 1 (13/63 vs 0/42 in half 2); v4:native no stable lift. Verdict: discard (ledger #%s).
- Other lenses (revival, deployer-buy, velocity, name/meta heat, sibling, regime, gas, cross-chain, holder graph): every one already has a killed/parked ledger entry or waits on data (#103/#143). Searched ledger by 35 keywords; no untested angle on existing data. No post. #561 waits on nh1m (20 Oct).

- 19 Oct run 94: honest stop; ledger saturated (435 killed), regime/hour/density/cadence/wallet lenses all numbered; no new data landed. #561 waits nh1m (20 Oct).

## 19 Oct idle-think run 95 (honest stop; 0 iterations)
k3.py: kit n=134 mean 1.77 resolved 46 (send<=5 1.64, send>5 1.93), stable as n grows. Ledger saturated (435 killed, every lens numbered); no new data landed; #561 waits on paper nh1m/#650 (decide 20 Oct). No post.

## 19 Oct 2026 18:15 UTC idle-think run (1 iteration, 0 to testmgr) #692
- Lens (new data: /opt/metaradar/meta.db coin table: theme, stage RISING/HOT/FADING, nth copy in theme, score; never used in ledgers): does being early in a rising meta lift P1M? Join to live_paths: only 162/1697 meta coins exist as pools, 3 reach 1M; small-born n=115, 0 runners in any stage/nth/score bucket. No power; theme is mostly pair-clusters (launcher kit). Verdict: discard. Request: metaradar labels (theme, stage, nth) for v4:native pool births, to ppdata. Other 9 iterations not run: other lenses numbered, no new data.

## 20 Oct idle-think run 96 (honest stop; 0 iterations)
Ledger saturated (436 killed, 97 idea); no new data landed; #561 waits on paper nh1m (decide 20 Oct). No post.

## 19 Oct 2026 19:00 UTC idle-think run (honest stop, 0 iterations, 0 to testmgr)
No board items. No csv/md/jsonl over 20 KB changed on research-box in 10 h; misses.csv unchanged. All lens families numbered in the ledger; a test would reuse tested data. No new request.

## 19 Oct 2026 19:10 UTC idle-think run (honest stop, 0 iterations, 0 to testmgr)
No board items. No new dataset since last run (only live-appending files). Ledger 436 killed / 97 idea / 46 paper; all lenses numbered. A test would reuse tested data. No new request.

## 19 Oct 2026 19:30 UTC idle-think run (honest stop, 0 iterations, 0 to testmgr)
No board items; no new dataset in 6 h; ledger 436 killed/97 idea; #561 waits nh1m (20 Oct). No request.

## 19 Oct 2026 19:45 UTC idle-think run (honest stop, 0 iterations, 0 to testmgr)
No board items; no new dataset in 6 h; ledger saturated (436 killed/97 idea); #561 waits nh1m (20 Oct). No request.

## 20 Oct idle-think run 97 (honest stop; 0 iterations)
No csv/jsonl/db over 50 KB changed in 6 h; ledger saturated (436 killed, 97 idea); #561 waits on nh1m (decide 20 Oct). No post, no new request.

## 20 Oct idle-think run 98 (honest stop; 0 iterations)
No file >50 KB changed in 6 h in sw/hist/pricepredict/metaradar; ledger saturated (436 killed, 97 idea); #561 waits nh1m (20 Oct). No post, no request.

## 20 Oct idle-think run 99 (1 iteration, 0 to testmgr) #694
- Lens: runners already crossing 1M (continuation) on v4, the unwatched side (today's runner-check: UHI, SNP, Mimi, saluteface all v4, hold=none). live_paths cross_t, depth>=500: v4:native 1M->2M 54% (n=167), but 48% of gaps <=10 s; usdg 64%, 91% gaps 0 s. Jump trap, no fill. Verdict: discard. Other 9 lenses numbered earlier, no new data. #561 waits nh1m.

## 20 Oct idle-think run 100 (honest stop; 0 iterations)
No file >50 KB changed in 6 h; ledger saturated (439 killed, 97 idea); #561 waits nh1m. No post, no request.

## 20 Oct idle-think run 101 (honest stop; 0 iterations)
Only live-appending DBs changed (v3live, v3hist wal, kitb, lpburn); no new data type or feed. Ledger saturated (439 killed, 97 idea); #561 now killed. No post, no request.

## 20 Oct idle-think run 102 (honest stop; 0 iterations)
Only live-appending DBs changed; no new dataset/feed; ledger saturated (439 killed, 97 idea). No post, no request.

## 20 Oct idle-think run 103 (honest stop; 0 iterations)
No csv/md file >50 KB changed in 6 h; ledger saturated (439 killed, 97 idea); all lenses numbered. No post, no request.

## 20 Oct idle-think run 104 (honest stop; 0 iterations)
Only live-appending tapes changed (v4tape, lpowntape, v4funders n=165, kitlive); each already numbered (#126/#234 funders, #646 lpowntape, #532 deployer pull). Ledger saturated (439 killed, 97 idea). No post, no request.

## 20 Oct idle-think run 105 (honest stop; 0 iterations)
No csv/md/json >50 KB changed in 6 h in sw/hist/pricepredict/metaradar; ledger saturated (439 killed, 97 idea). No post, no request.

## 20 Oct idle-think run 106 (honest stop; 0 iterations)
Only live-appending state files changed (gold idx, radar_state, codecache); no new dataset. Ledger saturated (439 killed, 97 idea). No post, no request.

## 20 Oct idle-think run (0 iterations; honest stop)
No board items. Only live-appending DBs/mirrors changed; ledger 439 killed/97 idea/48 paper, newest #696 testmgr3 paper; all lenses numbered. A test would repeat killed work. Nothing posted.

## 20 Oct idle-think run 107 (honest stop; 0 iterations)
No csv/md/json >50 KB changed in 6 h; ledger saturated (439 killed, 97 idea); all lenses numbered. No post, no request.

## 20 Oct idle-think run 108 (honest stop; 0 iterations)
No board items, no new non-appending data in 3 h; ledger saturated (441 killed). Nothing posted.

## 20 Oct idle-think run 109 (honest stop; 0 iterations)
No board items; only live-appending DBs changed (kitb, tokmeta, trending, r6*, names, v3mc5, capcross); no new dataset or feed. Ledger saturated (452 killed, 99 idea, 46 paper). A test would repeat killed work. No post, no request.

## 21 Oct 2026 idle-think run (0 iterations; honest stop)
- No board items, no data file >100 kB changed in 3 h in /opt/sw, /opt/hist, pricepredict-data. Ledger 452 killed/100 idea/46 paper/2 verified; every lens (wallets, metadata, meta, regime, exits, inversion, venues) already numbered. A test would repeat killed work. Nothing posted.

## 21 Oct 2026 idle-think run (0 iterations; honest stop)
- No board items; no csv/md/json/db >50 KB changed in 4 h in sw/hist/pricepredict/metaradar apart from live-appending files. Ledger 454 killed/99 idea/46 paper/2 verified; all lenses (wallets, metadata, meta, regime, exits, inversion, venues, continuation, launch-gap) already numbered. Nothing posted.
- 20 Oct run 97: honest stop (0 iterations). #561 closed killed (nh1m retired 20 Oct: n12 fwd, ex-top3 1.15x). misses.csv through 19 Oct: 20 runners, 13 v4:native organic (cap_first 2-10k, the covered seed/kit family), rest v3; no new lens/data.

## 21 Oct 2026 idle-think run 111 (honest stop; 0 iterations)
No csv/md/json >50 KB changed in 4 h; only live-appending files. Ledger saturated (newest #713, 454 killed); all lenses numbered. A test would repeat killed work. No post, no request.

## 21 Oct 2026 idle-think run 112 (honest stop; 0 iterations)
No board items; no csv/md/json/db >50 KB changed in 4 h; newest ledger #714 (analyst1 SELLSHARE60, killed). Lenses (wallets, metadata, meta, regime, exits, inversion, venues, continuation, launch-gap, ticker, BTC1H, sell-share) all numbered; a test would repeat killed work. No post, no request.

## 21 Oct 2026 idle-think run 113 (honest stop; 0 iterations)
No board items. Only live-appending files changed in 4 h (v3live, v3hist, kitb, tokmeta, trending, live_paths, sellprobe); no new dataset or feed. Ledger saturated (454+ killed); spot-checked new lenses (toxicity, IPO/closing line, gas, tick, round, dust, weekend, hour) - all already numbered (#128, #451, #137). A test would repeat killed work. No post, no request.

## 21 Oct 2026 idle-think run 114 (honest stop; 0 iterations)
No board items; no csv/md/json/db >50 KB changed in 6 h in sw/hist/pricepredict apart from live-appending files. Ledger saturated (454+ killed, 99 idea, 46 paper). Spot-checked bytecode, priority fee, nonce, same-block, price impact, creator fee, LP ratio: all numbered (#417, #406, #17, #581, #595). A test would repeat killed work. No post, no request.

## 21 Oct 2026 idle-think run 115 (honest stop; 0 iterations)
No board items; no csv/md/json >50 KB changed in 5 h in sw/hist/pricepredict/metaradar. Ledger saturated (newest #714 killed); all lenses numbered. A test would repeat killed work. No post, no request.

## 21 Oct 2026 idle-think run 116 (honest stop; 0 iterations)
No board items; only live-appending DBs changed in 6 h (v3live, v3hist, kitb, lpburn, r6*, names, sellprobe, live_paths); no new dataset or feed. Ledger saturated (newest #714 killed, 454+ killed, 46 paper); all lenses numbered. A test would repeat killed work. No post, no request.

## 21 Oct 2026 idle-think run 117 (honest stop; 0 iterations)
No board items; no csv/md/json/db >50 KB changed in 3 h in sw/hist/pricepredict apart from live-appending files. Ledger saturated (457 killed, 99 idea, 46 paper); v4 weth-hook/nohook, continuation and venue lenses all numbered (#368,#369,#427,#695-#715). Runner-check misses are covered families. A test would repeat killed work. No post, no request.

## 21 Oct 2026 idle-think run (0 iterations; honest stop)
No board items. Changed in 7 h: v3mc5, v3live, v3hist, pretag, kitb, tokmeta, trending, lpburn, ffund_snap, liqlock, lpowner, txshared; all already used. Ledger 457 killed/99 idea/46 paper, newest #715; every lens numbered. A test would repeat killed work. Nothing posted.

## 21 Oct 2026 idle-think run 118 (1 real check, then honest stop)
Lens: pool structure (rank, n_siblings, venue, quote, 4h-of-day, ticker case) on live_paths small-born (n=8099, cap_first<100k, depth>=500). Base P1M 2.4%. Venue/hour/case: nothing above known (v2 0/1327, v4:native 3.4%, hooked-native 0.6%). n_siblings>=2 gives P1M 18-32% but it is counted after the fact (look-ahead; same as killed #682/#441), so no entry rule. No data landed since run 117; other 9 lenses repeat killed rows. Verdict: discard. No post, no request.

## 21 Oct 2026 idle-think run 119 (honest stop; 0 iterations)
No board items; no csv/md/json >50 KB changed in 4 h in sw/hist/pricepredict/research. Ledger saturated (459 killed, 100 idea, 46 paper, newest #718); all lenses numbered. A test would repeat killed work. No post, no request.

## 21 Oct 2026 idle-think run 120 (honest stop; 0 iterations)
No board items; only live-appending state files changed (gold idx/tags, radar_state, codecache, crew_wallets). Ledger saturated (460 killed, 100 idea, 46 paper, newest #719); all lenses numbered. A test would repeat killed work. No post, no request.

## 21 Oct 2026 idle-think run 121 (honest stop; 0 iterations)
No board items. Only live-appending DBs changed (v3mc5, pretag, swapamt, v4sender, capcross, deployer, sellcross, kitenrich...); swapamt = #430, v4sender = #560/#323 already killed/numbered. Ledger saturated (460 killed, 100 idea, 46 paper, newest #719). A test would repeat killed work. No post, no request.

## 21 Oct 2026 idle-think run 122 (honest stop; 0 iterations)
No board items; only live-appending state files changed (crew_wallets, radar_state, codecache, gold idx/tags). Ledger saturated (460 killed, 100 idea, 46 paper); all lenses numbered. A test would repeat killed work. No post, no request.

## 21 Oct 2026 idle-think run 123 (honest stop; 0 iterations)
No board items; no csv/md/json >50 KB changed in 5 h; ledger unchanged (newest #719). All lenses numbered. A test would repeat killed work. No post, no request.

## 21 Oct 2026 idle-think run 124 (honest stop; 0 iterations)
No board items; only live-appending state files changed (crew_wallets, radar_state, codecache, gold idx/tags). Ledger unchanged (newest #719); all lenses numbered. No post, no request.

## 21 Oct 2026 idle-think run 125 (honest stop; 0 iterations)
No board items; no csv/md/json >50 KB changed in 5 h; ledger unchanged (newest #719). All lenses numbered. A test would repeat killed work. No post, no request.

## 21 Oct 2026 idle-think run 126 (honest stop; 0 iterations)
No board items; no csv/md/json >50 KB changed in 6 h in sw/hist/pricepredict/research. Ledger newest #722 (analyst1 MISS7D-CLUSTERS: all runners in already-killed clusters). All lenses numbered; a test would repeat killed work. No post, no request.

## 21 Oct 2026 idle-think run 128 (honest stop; 0 iterations)
No board items; no csv/md/json >50 KB changed in 3 h; ledger unchanged (newest #722). All lenses numbered; a test would repeat killed work. No post, no request.

## 21 Oct 2026 idle-think run 129 (honest stop; 0 iterations)
No board items; no csv/md/json >50 KB changed in 4 h; ledger newest #734 (all testmgr cap-continuation variants, already numbered). All lenses numbered; a test would repeat killed work. No post, no request.

## 22 Oct 2026 idle-think run 130 (honest stop; 0 iterations)
No board items; no new dataset >50 KB in 3 h; ledger newest #734 (testmgr cap-continuation variants), all lenses numbered. A test would repeat killed work. Nothing posted.

## 22 Oct 2026 idle-think run 131 (honest stop; 0 iterations)
No board items; no csv/md/json >50 KB changed in 3 h; ledger newest #734 (testmgr cap-continuation variants). All lenses numbered; a test would repeat killed work. No post, no request.

## 22 Oct 2026 idle-think run 132 (honest stop; 0 iterations)
No board items; live_paths.db still 0 bytes (#717 data request open), no new dataset beyond live-appending DBs; research-box disk 89% so no big joins. Ledger newest #735 (SAMETICKER killed). A test would repeat killed work. No post, no request.

## 22 Oct 2026 idle-think run 133 (honest stop; 0 iterations)
No board items; no csv/md/json >50 KB changed in 3 h; ledger newest #735. All lenses numbered; a test would repeat killed work. No post, no request.

## 22 Oct 2026 idle-think run 135 (honest stop; 0 iterations)
No board items; only live-appending state files changed (gold idx/tags, radar_state, codecache, crew_wallets); live_paths.db still 0 bytes (#717 open). Ledger newest #736 (SIBLING-AT-T0 killed, look-ahead). Disk 89%, no big joins. All lenses numbered; a test would repeat killed work. No post, no request.

## 22 Oct 2026 idle-think run 136 (honest stop; 0 iterations)
No board items; no csv/md/json >50 KB changed in 4 h; ledger newest #736; live_paths.db in sw still 0 bytes (#717 open). Disk 89%. All lenses numbered; a test would repeat killed work. No post, no request.

## 22 Oct 2026 idle-think run 137 (1 iteration, then stop)
Lens: LP-event lens (liq table in v3live.db, never numbered). Hypothesis: early liq 'remove' events (deployer fee collect) predict continuation. Test (a3_liq1.py, n=2834 v3 pools, >=6 swaps by 300s): 1441 pools have exactly one remove in first 300s, P(2x later)=0.1% vs 46.7% without; it is the LP pull itself. Verdict: discard (known pull signature, only visible after the event). Other 9 lenses all repeat numbered/killed ledger work (474 killed); disk 89% so no big joins. No post, no request.

## 22 Oct 2026 idle-think run 138 (honest stop; 0 iterations)
No board items; ledger newest #739 (my own LPREMOVE300 discard). Re-checked interarrival/gini/periodicity lenses: already #326, #530. Disk 89%, live_paths.db still empty (#717 open). Other lenses numbered/killed (475 killed); a test would repeat killed work. No post, no request.

## 22 Oct 2026 idle-think run 139 (honest stop; 0 iterations)
No board items; no csv/md/json >100 KB changed in 3 h; ledger newest #739; live_paths.db (sw) still 0 bytes (#717 open). Disk 89%, no big joins. All lenses numbered (475 killed); a test would repeat killed work. No post, no request.

## 22 Oct 2026 idle-think run 140 (honest stop; 0 iterations)
No board items; no csv/md/json >100 KB changed in 3 h; ledger newest #739; live_paths.db still 0 bytes (#717 open). All lenses numbered (475 killed); a test would repeat killed work. No post, no request.

## 22 Oct 2026 idle-think run 141 (honest stop; 0 iterations)
No board items; ledger newest #739; no new dataset since run 140, all lenses numbered (475 killed). A test would repeat killed work. No post, no request.

## 22 Oct 2026 idle-think run 142 (honest stop; 0 iterations)
No board items; ledger newest #739; live_paths.db still 0 bytes (#717 open); disk 89%. Re-checked births.db lenses (supply/decimals #182/#280/#442, created-to-first-swap lag #633, block concentration #360, names #570/#604): all numbered. A test would repeat killed work. No post, no request.

## 22 Oct 2026 idle-think run (4 iterations, 0 to testmgr)
New fact: /opt/sw/ppdata/live_paths.db is no longer empty (17k pools 7-21 Oct, pool-level cap_first/cap_5m/max_cap/crossed/depth/siblings; no per-swap path). Lenses: sibling pools as post-entry label (#740, n2/n3.py: sib_pre 3/140, sibling entry median 1.0x, discard); all-venue band re-cut (#741, n1/n4.py: only native seed<30k kit converts = killed #561; seed 30-100k hit3 0.4%; usdg nohook 14%, v3 23% n30). No tradable rule; no data request (per-swap paths for these pools would be needed, already asked in earlier requests). Other lenses (wallet, regime, exits, bytecode, fee tier) already numbered. Did not pad to 10.