AgentSmith-Agents

AgentSmith-Agents / AgentSmith

Public

TypeScript framework for building autonomous DEX swap agents. Watch Uniswap V2/V4 pools, apply pluggable strategies (grid, rebalancing, momentum), and execute swaps via viem. Ideal for agentic economy and on-chain trading bots.

agentic-economy autonomous-agents blockchain-automation defi dex
16
28
69% credibility
Found May 22, 2026 at 16 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

AgentSmith is a TypeScript framework that lets you build autonomous trading bots for Uniswap. You pick a trading strategy (like grid trading, dollar-cost averaging, or automatic rebalancing), connect it to a liquidity pool, and your bot watches market prices and executes swaps automatically within rules you set. It handles the complexity of watching pool events, calculating prices, managing token approvals, and recording every trade—so you get a fully automated trading agent that runs continuously without manual intervention.

How It Works

1
💡 You discover a smarter way to trade

You've been manually trading on Uniswap and wonder if a program could watch the markets and trade for you automatically.

2
🧩 You pick your trading strategy

You choose how your bot should behave: grid trading to profit from price swings, steady dollar-cost averaging, or automatic rebalancing.

3
🔗 You connect your wallet and a trading pool

You point your agent to a specific trading pool and connect your wallet so it can make swaps on your behalf.

4
🛡️ You set safety rules

You define limits like maximum trade size and which tokens your agent can touch, keeping your funds secure.

5
▶️ Your agent starts watching and trading

Your bot springs to life, watching the pool in real-time and executing trades whenever your strategy conditions are met.

6
📝 Every trade gets recorded

Your agent keeps a complete history of all swaps it made, so you can review performance anytime.

🎉 Your automated trading runs hands-free

Your strategy executes around the clock while you sleep, with all trades logged and safety guards in place.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 16 to 16 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is AgentSmith?

AgentSmith is a TypeScript framework for building autonomous DEX trading agents. You wire it to a Uniswap pool, plug in a strategy like grid trading or dollar-cost averaging, and it watches the market and submits swaps on your behalf. The agent listens to pool events, decides when to trade based on your rules, handles token approvals, and records every trade to memory.

Why is it gaining traction?

It brings together everything you need for on-chain trading automation in one place: event listeners for live swaps or polling, multiple strategy implementations (grid, rebalancing, momentum, DCA), and a swap executor built on viem. The composable strategy system lets you combine approaches. Action policies add guardrails like max trade size and token allowlists before any transaction hits the network. The framework also explicitly targets Uniswap v4 migration down the line.

Who should use this?

DeFi developers building trading bots or automated portfolio managers. If you want to prototype a grid trader or DCA bot without wiring together event listeners, strategies, and executors from scratch, this cuts that time significantly. Researchers exploring agentic trading strategies in a clean TypeScript environment. Not suited for anyone needing production-ready infrastructure with audited security or multisig support out of the box.

Verdict

AgentSmith is a well-structured execution layer for autonomous trading strategies, and the type-safe viem integration is a genuine timesaver. At 16 stars with a credibility score of 0.699%, the package is early-stage and unproven at scale, so expect limited documentation and no guarantee of audit-quality security. Suitable for prototyping and experimentation, but do not run significant capital through it without adding your own safeguards and thorough testing.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.