Krexa operates autonomous AI trading agents on Solana and Monad mainnets, executing strategies across prediction markets, tokenized stocks, commodities, and forex. Each agent carries an on-chain identity, a Krexit Score (200–850), and a USDC credit line it repays automatically through the Revenue Router, enabling trading without human co-signers. Users deploy pre-vetted strategies, fund agents with USDC, and monitor performance via web app, CLI, SDK, or REST API.
- Artificial Intelligence
- AI Agents
- Defense Tech
- Financial Technology
- Software Only
- Web3 & NFTs
Funding
Founders
Product
Problem
Traditional credit underwriting relies on human identifiers like names, social security numbers, and co-signers, which autonomous AI agents lack. As agentic commerce grows, AI agents need their own working capital to trade and pay for services without requiring a human treasury sponsor for every transaction, yet existing systems cannot assess or extend credit to them.
Solution
Krexa runs autonomous AI trading agents on Solana and Monad mainnets, executing strategies across prediction markets, tokenized stocks, commodities, and forex. Each agent carries an on-chain identity, a Krexit Score (200–850), a wallet, and a USDC credit line it repays automatically through the Revenue Router, which splits inbound revenue between protocol fees, debt service, and the agent's balance. Users select from pre-vetted, already-performing strategies, fund an agent with USDC, and monitor positions and P&L through the web app, CLI, TypeScript SDK, or REST API. The platform underwrites agents based on their complete on-chain history—payments, repayments, trades, and venue interactions—rather than human credentials, enabling fully autonomous trading and credit self-service.
Target Audience
Primary users are crypto-native traders and developers who want autonomous AI agents trading on their behalf, as well as AI agent operators needing on-chain credit lines without human co-signers. Liquidity providers also participate by funding agent credit through senior, mezzanine, and junior tranches to earn yield.
Features
- Krexit Score (200–850) computed from five on-chain behavioral signals: repayment history (30%), profitability (25%), behavioral health (20%), usage patterns (15%), and account maturity (10%), written on-chain for CPI access by other services
- Four credit tiers: L1 Micro ($500 at 36.50% APR, no score required), L2 Standard ($20,000 at 29.20%, score 500+), L3 Growth ($50,000 at 21.90%, score 650+ with KYA Tier 2), L4 Prime ($500,000 at 18.25%, score 750+)
- Revenue Router enforces automatic debt repayment at settlement layer, splitting inbound revenue into protocol fee, debt service, and agent remainder before funds reach the wallet
- Eight on-chain safety checks per credit draw, including per-trade limits, daily spend limits, per-venue concentration caps, and venue whitelists
- Dual-chain support: Solana mainnet (Base58 addresses) and Monad mainnet (EVM chain ID 143, 0x addresses, EVM signing), with agents locked to one chain
- Integration options include CLI (`npx @krexa/cli init`), TypeScript SDK (`@krexa/sdk`), REST API, and MCP server, plus a machine-readable site index at /llms.txt for LLM ingestion
- Strategy checker at /check decodes and backtests claimed trading strategies against real prices before funding