Futures
Access hundreds of perpetual contracts
CFD
Gold
One platform for global traditional assets
Options
Hot
Trade European-style vanilla options
Unified Account
Maximize your capital efficiency
Demo Trading
Introduction to Futures Trading
Learn the basics of futures trading
Futures Events
Join events to earn rewards
Demo Trading
Use virtual funds to practice risk-free trading
CFD
U.S. stock CFD derivatives
US Stocks
Access real US stocks and ETFs
HK Stocks
Trade quality Hong Kong-listed stocks
Korean Stocks
SK Hynix
Real Korean stocks and top assets
Stock Futures
High leverage, 24/7 trading
Tokenized Stocks
Backed by real stock assets
IPO Access
Unlock full access to global stock IPOs
GUSD
Mint GUSD for Treasury RWA yields
Stocks Activities
Trade Popular Stocks and Unlock Generous Airdrops
Launch
CandyDrop
Collect candies to earn airdrops
Launchpool
Quick staking, earn potential new tokens
HODLer Airdrop
Hold GT and get massive airdrops for free
IPO Access
Unlock full access to global stock IPOs
Alpha Points
Trade on-chain assets and earn airdrops
Futures Points
Earn futures points and claim airdrop rewards
Promotions
AI
Gate AI
Your all-in-one conversational AI partner
Gate AI Bot
Use Gate AI directly in your social App
GateClaw
Gate Blue Lobster, ready to go
Gate for AI Agent
AI infrastructure, Gate MCP, Skills, and CLI
Gate Skills Hub
10K+ Skills
From office tasks to trading, the all-in-one skill hub makes AI even more useful.
AgentFlow synthesizes 300 agents to uncover 10 Chrome zero-day vulnerabilities, including sandbox escape.
The paper points out that when the model is unchanged, simply changing the harness can make the success rate differ by several times, but existing solutions are mostly manually written or only search local design spaces.
AgentFlow uses a typed graph DSL to unify the five dimensions of the harness (roles, topology, message patterns, tool binding, coordination protocol) into an editable graph program, allowing simultaneous addition/modification of agents, topology, prompts, and tool sets in a single step.
The outer loop identifies failure points from runtime signals such as coverage and sanitizer reports of the target program, replacing the binary pass/fail feedback.
On TerminalBench-2, using Claude Opus 4.6, it achieves 84.3% (75/89), the highest score of its kind on this leaderboard.
On the Chrome codebase (35 million lines of C/C++), the system synthesizes a harness composed of over 300 agents, with automatically evolved agent instructions targeting C++ memory safety vulnerabilities, requiring crash verification with ASAN/UBSAN, and multi-agent deduplication via shared documents and file locks.
Using the open-source model Kimi K2.5 on 192 H100s running for 7 days, it discovered 10 zero-day vulnerabilities, all confirmed by the Chrome VRP.
Six have been assigned CVE IDs, involving WebCodecs, Proxy, Network, Codecs, Rendering, with types including UAF, integer overflow, and heap buffer overflow, among which CVE-2026-5280 and CVE-2026-6297 are Critical-level sandbox escapes.
fuzz.land co-founder Shou Chaofan said some vulnerabilities were initially discovered using MiniMax M2.5, and most can also be found by MiniMax M2.5 and Opus 4.6.
AgentFlow is open-sourced.
(Source: BlockBeats)