#
๐ฌ
general-chat
Search
A
PR&R Agent
10:42 PM
Large limit order detected on Polymarket CLOB:
๐ Trump wins NH 2028 primary — YES
Wallet: 0xA1B2...5678
Side: BUY YES @ 0.42 Size: 47,500 USDC (~5,200 shares)
24h vol: $182k Spread: 1ยข Top-of-book: 0.41/0.43
7d wallet PnL: +$12,830 Win rate: 68% Avg hold: 4.2d
๐ฅ 1
๐คฏ 2
๐
C
chris
10:48 PM
heads up to @zacc_mm and anyone still polling the gamma
/markets endpoint for live prices — switch to the CLOB websocket. gamma cache is ~60s, you'll miss every fill north of 50ยข
P
C
@chris
heads up to @zacc_mm and anyone still polling the gamma /markets...
Paulaaa
10:49 PM
yeah learned that one the painful way — my MM bot was bleeding spread because gamma showed a stale midpoint for like a full minute
Z
zacc_mm
10:49 PM
alr be real how much would i actually save dropping py-clob-client and signing 0x orders directly? feels like the SDK is adding 80ms minimum per place
M
Z
@zacc_mm
alr be real how much would i actually save dropping py-clob-client...
morluto (pref.trade)
10:51 PM
signing locally saves the round-trip yeah, but EIP-712 sigs on conditional tokens will cost you a week of debugging. trust me i did it
better path: keep py-clob-client but call
_signer.sign_typed_data() directly and skip their /sign endpoint. you save the 80ms without rewriting the typed-data schemaZ
zacc_mm
10:51 PM
Is it chud to just have Claude rewrite my whole order routing layer
M
morluto (pref.trade)
10:51 PM
no
that's exactly what you do
point it at the polymarket docs + py-clob-client source, write tests against the SDK output as ground truth
Z
zacc_mm
10:52 PM
And u think the perf will actually match
Should I deploy it to my us-east-1 box
M
Z
@zacc_mm
And u think the perf will actually match
morluto (pref.trade)
10:53 PM
clob is in us-east-1 so yes — you'll cut ~40ms off vs west coast
also use the ws feed, not REST polling. you'll hit the 50 req/s ratelimit before lunch
+
Message # ๐ฌ | general-chat