# V2DEC9 $5 round-trip sell test: plan (exec, 12 Oct)
No real trades by Claude and no live switch. Two parts.

**Part 1: eth_call sim (Claude can do, no money).** For the next kit pair (factory 0x8bce, seed 1.0 ETH, token reserve 1e18):
1. At +10 s read getReserves on the pair (free rpcpool).
2. eth_call with state override: give a fresh test address 0.01 ETH and call the v2 router `swapExactETHForTokensSupportingFeeOnTransferTokens` (~$5), then the token approve and `swapExactTokensForETHSupportingFeeOnTransferTokens` for the whole balance, in one batched call (Multicall-style helper contract via state override code). Pass if the sell does not revert and ETH out >= 0.9 x ETH in minus the pool-curve cost.
3. Repeat at +60 s, +200 s and +340 s to check the sell hook does not close later (the 0x90bfb865 hook on other tokens reverted only the sell, so the sell leg at each time is the point).
4. Note: fm_sim.py on firstmin is v3-only (Sim.sol); v2 needs a small new helper. Build as /opt/crew/work/exec/dec9_sim.py when a pair appears (about 4 a day), triggered by the testmgr signal file; output one line per pair to fw/exec-dec9-sims.csv.

**Part 2: real $5 round trip (Ben only).** Only after 3 consecutive eth_call passes. Ben gets the exact CA, amount ($5), buy at +10 s, sell at +350 s in one message when it is ready; Claude roles never place it.

Status: no pair exists yet, so no result. Next kit pair is expected within hours. Item stays a plan; sim runs when testmgr posts the pair.