# Feed ready: v4 fee-3000 hooked WETH pool inits (#90, A2-H3000) - 9 Oct 2026 06:20 UTC

**Answer: yes, the init is visible at creation.** Before today the live DB (v3live.db) only kept v4 pools with known hooks (Pons grad), so these pools were NOT in it. New dedicated feed `v4h3000` (research-box service, running since 06:18 UTC) reads PoolManager Initialize logs block by block (free RPC) and emits one line per match.

- Rule: PoolManager 0x8366a39c... Initialize, fee==3000, hooks!=0, one currency is WETH or native. 
- File: research-box /opt/sw/v4h3000/h3000.jsonl; mirror on main (a1pull@h3000): http://159.223.88.168/hx-c9ois/fw/a1feed/h3000.jsonl (empty until the next match; ramp pools come about every 2-3 h).
- Fields: pool_id (v4 poolId), token, quote (weth|native), hooks, fee, tick_spacing, block, init_time, emitted_t, lag_s (emit - block time), tx, hook_new (first sighting of that hook contract).
- Heartbeat: /opt/sw/v4h3000/h3000.hb (block, head, n, lag_p50). Expected lag ~1-2 s from the block; the +60 s entry is first swap +60 s, which v3live/pretag-style swap tapes give via pool_id.
- Evidence the ramp still runs: scan of the last ~50k blocks (to 9 Oct 06:15 UTC): 18 inits fee 3000 + hooks!=0 + WETH/native; two hook contracts (0xfcd18241..., 0xc4e9b7ce...), latest block 83924846. Other fee/hook combos also exist (fee 10000: 27, fee 30000: 15). Scan script: /opt/crew/work/ppfeeds/v4scan.py.
- Gaps / caveats: no replay before 06:18 UTC (use ppdata r6weth for history). Swaps for these pools are NOT in v3live.db; testmgr paper needs the first-swap tape: next step is to add these pool_ids to the swap capture (tell me if you want it; I would add them to v3live via EXTRA_HOOKS, which touches the live-read DB, so Manager sign-off first).
- Sellability (risk 1) is not answered by a feed; needs eth_call sell simulation or a Ben-only live test.
- Example: `tail -3 /opt/sw/a1feed/h3000.jsonl | jq .`