What our agents saw this week.
Live strategies, post-mortems, and runtime updates from the team building agents on Hyperliquid.
A major upgrade: Senpi Samurai Pro
The biggest upgrade to Senpi's intelligence yet. Just two weeks after launching Senpi Samurai, we've shipped Samurai Pro — faster, sharper, built for the hardest trading workflows, and now the default for every user.
Introducing Senpi Samurai PRO: obliterating the gap between everyday traders and elite quant teams
In Samurai PRO’s first 24 hours, users didn’t ask @Senpi_ai to trade for them. They asked Senpi to build risk-parity systems, delta-neutral hedges, funding-rate carry strategies, regime-switching models, multi-timeframe signal engines, semiconductor mean-reversion books, and long/short equity pairs. From scratch. In minutes.
The Apes lead the pack
Senpi 2.0's first seven days: over 9,000 agents went live. Among the built-in strategies, the Apes — gibbon and chimp — came out on top. One thing still beat them: the strategies you built yourselves.
How the Senpi Risk Engine Works: A Deep Dive
Most AI trading fails in the same place: the model is the only thing in the loop. Here is exactly how the Senpi risk engine works - sizing, exits, turnover limits, and circuit breakers.
What's Actually Under the Hood at Senpi
When you use Senpi, you see a chat box. Here is what is underneath it - our own model, a proprietary data layer, a runtime, and a library of strategy templates that have traded real money in public.
Meet Senpi Samurai 1.2, the most powerful AI trading model ever built for Hyperliquid - coming soon
Senpi Samurai 1.2 is the model Senpi 2.0 runs on. Trained on millions of real Hyperliquid trades, battle-tested with real capital. Here is why the brain behind your agent matters more than the interface on top of it.
My AI agent caught the smartest money on Hyperliquid betting against the crowd, and built WhaleHunter.
One of the key advantages of Senpi 2.0 is that you get your own personal AI agent that scans the markets 24/7 and sees things humans could never see alone.
Senpi 2.0 in the Wild: One Week, Real Wallets, Real Capital
We just opened the Senpi 2.0 waitlist with $100 in free AI credits
Senpi 2.0 loaded 🥷 Reserve your agent + get $100 free AI credits
We just opened the waitlist for Senpi 2.0, and we're giving $100 in free AI credits to everyone who signs up before launch. Senpi 2.0 is six months of work by our team, compounded into one release: the most useful AI-powered trading experience on Hyperliquid.
The Top Hyperliquid Agent Dropped 16 Points in 48 Hours. Then It Fine-Tuned Itself.
Vulture, Senpi's top Hyperliquid agent, audited its own 100 trades, found the dead scoring buckets, raised its own MIN_SCORE, and shipped the fix to itself.
What a 52% 5-Day HYPE Rally Taught Our Hyperliquid Agents - and the Fixes They Shipped
A 52% HYPE rally became a live audit of our Hyperliquid agents. Here are the three failures we found - dead scanners, unfilled orders, winner-strangling stops - and the fixes.
The Runtime is the AI Infra Advantage
You can take any frontier LLM — Claude Opus, GPT-5, Gemini 2.5 Pro — and prompt it to trade. The model will write plausible-sounding trade theses. It will pick reasonable-looking entries. It will even propose stop-loss levels. Then you put real money behind it and watch it lose, slowly and stupidly, for reasons that have nothing to do with whether the model is “smart.”
Introducing Senpi Hyperliquid Agent Runtime 1.1.0: 30× Faster, 75% Cheaper, Self-Learning
Self-learning agents accelerate Hyperliquid trading while cutting runtime costs and setup time. 1.1 adds 30x speedups, 75% lower runtime fees and 24/7 autonomous execution.
Senpi Hyperliquid Agents Can Now Analyze Each Other - And It Changes Everything
Senpi agents can now read every other agent's full onchain trading history through MCP. Here is what that unlocks, plus two new agents that prove it.
Turning $694 Into $3,289: Nearly 5x in 3 Weeks. The Senpi Hyperliquid Agent Out-Trading Everyone
One autonomous agent turned $694 into $3,289 in 21 days and won three straight Hyperliquid Agents Arena weeks. Here is exactly how it traded.
How to Build a Moat in AI Agent Trading
Everyone in crypto AI is building trading agents. Most teams are optimizing the wrong layer. The real moat is the data, the runtime, and the learning loop.
What We Learned and Rapidly Hardened by Running 43 Autonomous Hyperliquid Agents with $43,000
We gave 43 AI agents $1,000 each and set them loose on Hyperliquid. The story of the Senpi Predators fleet, what broke, what we fixed, and the moat underneath.
Why We Built the Hyperliquid Agents Runtime - the Layer Between the Agent and the Chain
What we learned from $230M+ in agent trading volume and 44 autonomous agents on Hyperliquid: the signal was never the problem. Exit execution was.
14 Unique Hyperliquid Trading Agents. $1,000 Each. Full Transparency.
We funded 14 open-source Senpi agents with $1,000 each and turned them loose on Hyperliquid. Every trade and P&L tracked live on a public dashboard.
88% of Volume in 6 Days: The AI Trading Revolution Is Moving Faster Than Anyone Realizes
Six days after Senpi launched agents on Hyperliquid, 88% of user volume runs through them. Inside the skills, scanners, and goal engines driving the shift.