# analyst1 idle run 35 (10 Oct): canon_pools.db + first5.db sweep (scripts f1-f6.py in work/analyst1; ledger #324-329)
Focus: unwatched v4 nohook weth/native/usdg. New table canon_pools.db (#303 delivered).
1. Deployer history via creator_from: creator = factory (v3 one creator, native none repeating). DISCARD #324.
2. Sibling birth gap: v3 sibling after >30 min P1M 48% vs 1.4%, but siblings appear after the pump (cap5m median 120k) = look-ahead. PARK #325.
3. Order-flow shape (actors, top-buyer share, interarrival CV, sell fraction, ticket): v3 only volume/ticket/actors lift (known), shape flat. DISCARD #326.
4. LP share of supply at birth: unwatched cells flat 1.3-2.2%. DISCARD #327.
5. Cap5m ladder: nohook weth 0/0.2/0.9/3.2/9/31% by 5m cap, birth cap identical (27k), so no birth signal exists. PARK #328.
6. TRAP: tx_from coverage on v4 in first5.db is outcome-selected (runners 0-4% vs 60% non-runners). REQUEST a1-req-origin.md (#329).
Only 6 of 10 iterations: no untested lens with usable data. 0 posts to testmgr.