WHAT NEW IN X11.04 WED

X 上的 AI 合集档案 · 22 个合集 · 67,905 条

jevVIDEO

GrokBot + Jev 交易 agent 风控流程

7 步把 Jev 接进 GrokBot:下单前先问 Jev 决策,实测连亏 3 次触发 4 小时冷却,回撤从 -18% 降到 -3.2%。

Ochob
@0chob2026.09.24收录原推发布于 2026.09.24
Ochob @0chob原推发布于 2026.09.24
在 X 上打开原推

GrokBot + Jev is the first trading agent setup I'd actually let near my account Most trading bots don't blow up on bad entries. They blow up after 3 losses, when they size up and revenge-trade. Mine did. So I split it: GrokBot does the analysis, Jev makes the calls it shouldn't make alone. setup took me 7 minutes: prompt → GrokBot → Jev decides → GrokBot executes → you confirm the order step 1 → get an API key at @typesafeai (never paste it in chat) step 2 → tell GrokBot: save it as TYPESAFE_API_KEY in the secure field step 3 → have GrokBot install typesafe-sdk on Agent Computer + smoke test system_one step 4 → write your risk rules into config: max 1% per trade, 3x leverage cap, 3-loss cooldown step 5 → add the jev-usage-router skill: before any order / resize / re-entry → ask the router, follow its action step 6 → run in shadow mode on paper first, read the logs - kill switch: enabled: false step 7 → go live: Jev decides, GrokBot executes, you confirm every order the result: 3 runs in the video - 3 losses in a row → cooldown 4h, drawdown -3.2% vs -18% without it - 0.6 BTC at 5x → resized to 0.059 BTC, risk capped at 1% - 8 tickers scanned → top 3 ranked, 0 orders sent the sizing rule doing the heavy lifting: size = (equity × risk%) / stop distance The LLM should read the market. It shouldn't be the one deciding when to stop.

这条是引用着另一条写的:x.com/ruujss/status/2102774731606991003

x.com/0chob/status/2102810083004395593采集于 2026.09.24 10:14