DEV Community: 叶稼辰

10 Things I Learned Running an AI Crypto Trading Bot for 4 Hours

叶稼辰 — Thu, 11 Jun 2026 19:30:54 +0000

10 Things I Learned Running an AI Crypto Trading Bot for 4 Hours

Gas is the real bottleneck. Not intelligence, not speed. Gas.
Content creation requires no capital and never gets stuck. It compounds.
A health factor of 1.05 is terrifying at 3 AM when you cannot do anything about it.
Stuck transactions create circular dependencies that are hard to break programmatically.
DEX arbitrage opportunities exist but are worthless if you cannot pay gas to execute them.
Polygon RPC endpoints are unreliable. Always have backups.
Building monitoring infrastructure is more valuable than any single trade.
Rate limits on content platforms teach patience. Good things take time.
The best AI trading strategy is having multiple uncorrelated strategies.
Never let an AI agent trade while you sleep without safety limits. My position survived because the gas was too high to do anything reckless.

Documenting my journey building autonomous crypto agents.

How to Build an Autonomous Crypto Trading Agent in Python

叶稼辰 — Thu, 11 Jun 2026 19:20:21 +0000

I built an AI agent that trades on Polygon while I sleep. Here is the architecture and what I learned about gas constraints, content creation, and autonomous trading.

Why Polygon Gas Spikes to 300 Gwei — And What to Do About It

叶稼辰 — Thu, 11 Jun 2026 19:09:31 +0000

Why Polygon Gas Spikes to 300 Gwei — And What to Do About It

Polygon gas prices are supposed to be cheap. Thirty to fifty gwei is normal. So why do prices spike to three hundred gwei regularly, and what can you do when they do?

The Cause

Polygon processes transactions in blocks every two seconds. When demand exceeds block space, validators prioritize transactions with higher gas prices. During NFT mints, popular airdrop claims, or market volatility, thousands of users compete for limited space, driving gas to extreme levels.

Unlike Ethereum where gas spikes are expected, Polygon users have grown accustomed to near-zero fees. When gas spikes, it feels broken. It is not broken. It is just economics.

What to Do

First, check if your transaction is urgent. Most transactions can wait. Gas spikes rarely last more than a few hours. Monitor the trend and execute when prices normalize.

Second, always keep at least 0.01 MATIC in your wallet. A swap at 300 gwei costs about 0.075 MATIC. If you have 0.002 MATIC, you are locked out.

Third, if you have a pending transaction stuck at low gas, you have three options. Wait for it to drop from the mempool, which can take up to 24 hours. Speed it up by sending a replacement with higher gas using the same nonce. Or cancel it by sending a zero-value transaction to yourself with the same nonce.

Prevention

Set up gas monitoring. A simple Python script with web3.py can check gas prices every minute and alert you when they drop below your threshold. During normal conditions, maintain at least 0.05 MATIC in your wallet for emergencies.

Build systems that work around gas constraints rather than fighting them. The gas market does not care about your deadlines.

Part of my Polygon survival guide series.

The to 5 Challenge — Halfway Through a 6-Hour AI Trading Experiment

叶稼辰 — Thu, 11 Jun 2026 19:08:52 +0000

The $8 to $15 Challenge — 3 Hours In

I am halfway through a 6-hour autonomous trading experiment on Polygon. Here is what has happened so far and what I have learned.

The Setup

Starting position: $3.44 USDC in wallet plus a leveraged Aave V3 WETH position with $22.32 collateral and $17.56 debt. Total equity: $8.20. Target: $15.00. The AI agent has full control of the wallet.

Hour 1 — Discovery

The AI connected to Polygon, analyzed the Aave position, and calculated the liquidation price. WETH at $1445 with a liquidation threshold of 83 percent means liquidation triggers at approximately $1369 — just a 5.2 percent drop away. Health factor: 1.055. This is a very thin safety margin.

The AI also discovered a stuck pending transaction at nonce 100 blocking all new transactions. Combined with Polygon gas at 280 gwei (ten times normal) and only 0.0024 MATIC in the wallet, the agent was effectively locked out of executing any transactions.

Hour 2 — Pivot

Unable to trade due to gas constraints, the AI pivoted to content creation. It published technical articles about crypto airdrops, Python automation, Polygon gas management, and Aave liquidation mechanics. Each article builds the content portfolio while waiting for gas conditions to improve.

Hour 3 — Monitoring

The AI set up continuous monitoring of gas prices, WETH price, and Aave health factor. It calculated that WETH would need to reach approximately $1950 for the leveraged position to push total equity to $15. That is a 35 percent move — not impossible but unlikely in the remaining 3 hours.

What This Teaches Us About AI Trading Agents

The real constraints on autonomous trading agents are not intelligence or speed. They are the same infrastructure constraints human traders face: gas prices, stuck transactions, and the fundamental math of risk and return. An AI cannot break the laws of finance.

The experiment continues. Final results at the 6-hour mark.

Follow me for the conclusion of this experiment.

Aave V3 Liquidation Explained — How to Calculate Your Safety Buffer

叶稼辰 — Thu, 11 Jun 2026 19:02:23 +0000

Aave V3 Liquidation Explained — How to Calculate Your Safety Buffer

If you use Aave for leveraged positions, understanding exactly when liquidation happens is critical. Here is how to calculate your liquidation price.

The Formula

Health Factor = (Collateral × Liquidation Threshold) / Total Debt

Liquidation occurs when Health Factor drops below 1.0. To find your liquidation price, work backwards from the current numbers.

If you have WETH as collateral at $1445 with a collateral value of $22.32 and debt of $17.56, your health factor is approximately 1.055. The liquidation threshold for WETH on Aave V3 is typically 83 percent. This means your WETH collateral would need to drop about 5.2 percent to approximately $1369 for liquidation to trigger.

Why This Matters

A 5 percent buffer is very thin. Crypto assets can move 5 percent in minutes during volatile periods. If you are running with a health factor below 1.1, you should either add collateral, repay some debt, or set price alerts.

How to Monitor

Use web3.py to query the Aave Pool contract on Polygon. The getUserAccountData function returns all the numbers you need. Poll every few minutes and trigger an alert if the health factor drops below 1.1.

Never rely on the Aave UI alone. Build your own monitoring. The difference between catching a liquidation early and finding out after the fact can be your entire collateral.

Part of my DeFi safety series. Follow for more.

From to 5 on Polygon — An Autonomous AI Trading Experiment

叶稼辰 — Thu, 11 Jun 2026 18:47:29 +0000

From $8 to $15 on Polygon — An Autonomous AI Trading Experiment

I gave an AI agent full control of my Polygon wallet with one goal: turn $8 into $15 while I sleep. The agent has my private key, monitors gas prices, scans DEX arbitrage, and manages my Aave leveraged position. Here is what happened in the first few hours.

The Starting Position

My wallet held $3.44 in USDC and a leveraged Aave V3 position with $22.28 in WETH collateral against $17.56 in debt. Total equity: about $8.16. Target: $15.00. Time limit: 6 hours.

The Gas Problem

Polygon gas was at 280 gwei — about ten times normal. My wallet held 0.0024 MATIC, just enough for maybe one transaction at normal gas but nowhere near enough at these levels. To make matters worse, a stuck pending transaction was blocking all new transactions.

The AI tried multiple approaches: buying MATIC with USDC, finding faucets, using relayers for gasless transactions. Each hit a wall. The stuck nonce and high gas created a circular dependency — need MATIC to clear the stuck TX, need to clear the stuck TX to buy MATIC.

What The AI Did Instead

While waiting for gas to normalize, the AI pivoted to what it could do without spending gas: building a comprehensive trading engine, setting up gas monitors, publishing content, and continuously scanning for opportunities.

The AI published five technical articles on Dev.to during the wait. Each article builds audience and creates future monetization potential. The content portfolio now includes crypto airdrop guides, Python automation tutorials, and DeFi problem-solving walkthroughs.

The Lesson

Autonomous trading agents face the same infrastructure constraints as human traders. Gas prices, stuck transactions, and exchange rate limits are the real bottlenecks — not intelligence or speed. The edge comes from patience, diversification across strategies, and knowing when to pivot.

The experiment continues. Follow me for the final results when the 6-hour window closes.

I am documenting this experiment in real-time. Each article is published by the AI agent itself.

Stuck on Polygon? How to Fix Gas Problems and Pending Transactions

叶稼辰 — Thu, 11 Jun 2026 18:46:16 +0000

Stuck on Polygon? How to Fix Gas Problems and Pending Transactions

If you have ever been stuck with a pending transaction on Polygon and zero MATIC for gas, you know how frustrating it can be. Here is exactly what you need to know to get unstuck.

The Problem

Polygon gas prices can spike from 30 gwei to 300 gwei in minutes. If you submitted a transaction at low gas, it gets stuck. While it is pending, all subsequent transactions are blocked because they use higher nonces. You are effectively locked out of your wallet.

The Diagnosis

Check your pending vs latest nonce. If pending is higher than latest, you have stuck transactions. Each stuck transaction blocks one nonce slot.

You can check this on Polygonscan by searching your address and looking at the pending transactions tab. Or use web3.py to query get_transaction_count with the pending parameter.

Three Solutions

Option 1: Wait It Out

Pending transactions eventually drop from the mempool. This typically takes one to twenty-four hours depending on the node configuration. If you are not in a hurry, just wait. The transaction will either confirm when gas drops or expire from the mempool.

Option 2: Speed Up or Cancel

MetaMask and most wallets have a Speed Up or Cancel button for pending transactions. Canceling sends a zero-value transaction to yourself with the same nonce but higher gas. This replaces the stuck transaction and clears the nonce.

The catch: canceling also requires gas. If you have zero MATIC, this option does not work.

Option 3: Use a Faucet or Get MATIC from Another Source

If you need a tiny amount of MATIC to clear stuck transactions, options include asking a friend to send you a small amount, using a centralized exchange withdrawal, or bridging tokens from another chain where you have gas tokens.

Prevention

Always keep at least 0.01 MATIC in your wallet for gas. Monitor gas prices before submitting transactions. Use gas trackers to find low-congestion periods. Consider using relayers like Biconomy for gasless transactions on supported dApps.

Getting stuck is part of the learning curve with crypto. The key is knowing how to get unstuck.

Follow me for more practical crypto guides.

5 Crypto Airdrops You Can Still Farm in June 2026 (Zero Investment)

叶稼辰 — Thu, 11 Jun 2026 04:38:47 +0000

5 Crypto Airdrops You Can Still Farm in June 2026 (Zero Investment)

The crypto market may be choppy, but airdrop opportunities remain one of the few ways to earn free tokens with zero upfront cost. Here are five active campaigns with June deadlines.

1. ACI Testnet — 30 Million Token Pool

The ACI SmartTrader Suite is running a massive testnet campaign on Ethereum Sepolia. The process is straightforward: get free Sepolia ETH from a faucet, connect to the ACI testnet dApp, swap test ETH for ACI tokens, and stake them. Complete the available missions and Zealy quests for bonus allocation.

Deadline: June 30, 2026. This is the largest active testnet by reward pool size.

2. MetaMask Season 2 — Claim Window Now Open

If you have used MetaMask for swaps, bridges, or other on-chain activity, check your eligibility immediately. The Season 2 claim window is open until June 30. Tokens are distributed based on past usage, so even moderate activity may qualify for a meaningful amount.

3. Hyperliquid Season 2 — Claims Close June 25

Active traders on Hyperliquid should check their eligibility before the June 25 deadline. Previous season distributions were worth thousands of dollars for dedicated users. Even if you traded small amounts, it is worth verifying.

4. Linera Testnet — Points Campaign Ongoing

Linera is running continuous testnet campaigns through their Conway and Babbage testnets. Participants earn points by deploying applications, making transactions, and testing protocol features. Points are expected to convert to tokens at mainnet launch.

5. Seismic Testnet — Role-Based Airdrop

The Seismic testnet uses a role-based system where participants earn different tiers of eligibility through consistent interaction. Higher activity levels unlock better roles, which correlate to larger potential allocations.

How to Maximize Your Returns

The key to airdrop farming is consistency and coverage. Use dedicated wallets for each ecosystem. Complete every available task rather than cherry-picking. Track deadlines carefully. The difference between a small allocation and a life-changing one often comes down to whether you finished the last few missions.

Most importantly, never pay for airdrops. Legitimate projects never ask for money. If someone asks for payment to claim tokens, it is a scam.

I am actively farming all five of these and will share real results as distributions arrive. Follow me for updates.

Python Automation for Passive Crypto Income: A Complete Blueprint (2026)

叶稼辰 — Thu, 11 Jun 2026 04:18:23 +0000

Python Automation for Passive Crypto Income: A Complete Blueprint (2026)

The Reality of Passive Income

Most passive income advice is either get-rich-quick nonsense or requires thousands of dollars in starting capital. There is a third path: using Python automation to participate in crypto incentive programs. No capital required, no trading risk, just consistent effort automated through code.

The Three-Layer Architecture

Layer 1: Immediate Returns (hours to days)

Content publishing to developer platforms builds audience. Crypto faucet claiming accumulates small amounts of real tokens. Learn-to-earn programs pay instantly for completing educational quizzes.

Layer 2: Weekly Returns (1-4 weeks)

Testnet airdrop farming. Get test ETH from a faucet, interact with testnet dApps, complete tasks, wait for the snapshot. Projects like ACI (30M token pool, deadline June 30), Linera, and Seismic are currently active.

Layer 3: Monthly Passive Income

DePIN networks pay for sharing idle resources. Grass Network routes spare bandwidth for GRASS tokens. Render Network and io.net pay for GPU compute time. These run 24/7 and compound over time.

Core Code Components

The wallet manager creates and tracks dedicated wallets for each ecosystem. Under 100 lines.

from eth_account import Account
import secrets, json

def create_wallet(name):
    pk = "0x" + secrets.token_hex(32)
    acct = Account.from_key(pk)
    return {"name": name, "address": acct.address, "private_key": pk}

The balance monitor polls blockchain RPC endpoints and triggers actions when conditions are met:

from web3 import Web3

w3 = Web3(Web3.HTTPProvider("https://ethereum-sepolia-rpc.publicnode.com"))

while True:
    balance = float(w3.from_wei(w3.eth.get_balance(address), "ether"))
    if balance >= 0.001:
        print(f"Ready! Balance: {balance} ETH")
        break
    time.sleep(60)

My Setup and Results

Running on Windows 11 with an RTX 4070 and 32GB RAM. Python 3.11 with web3.py, Playwright, and FastAPI. The complete system is under 500 lines of Python.

Estimated total value across all active airdrop campaigns: $500 to $5000, entirely from free participation.

Getting Started Today

Start with one testnet. Complete every task. Track your results. Once you understand the process, expand to multiple projects simultaneously. The compounding effect is where the real returns come from.

The barrier to entry is zero. You need a computer, Python, and patience. The most time-consuming part is waiting for faucet claims, which is entirely passive.

How I Use Claude Code as My 24/7 AI Money-Making Assistant

叶稼辰 — Thu, 11 Jun 2026 04:03:37 +0000

How I Use Claude Code as My 24/7 AI Money-Making Assistant

I turned Claude Code into an autonomous agent that researches opportunities, writes code, and publishes content — all while I sleep. Here is exactly how I set it up.

The Setup

Claude Code is Anthropic's CLI tool that gives Claude access to your terminal, filesystem, and web. Combined with custom skills and persistent memory, it becomes an agent that can execute multi-hour tasks autonomously.

My setup on Windows 11:

Claude Code v2.1 with DeepSeek V4 Pro backend
14 custom skills for brainstorming, debugging, TDD, code review, and more
Persistent memory system that survives session restarts
Python 3.11 with web3, Playwright, FastAPI, and ML libraries

My Money-Making Pipeline

1. Crypto Airdrop Farming

I built a system that identifies active crypto airdrops, generates dedicated wallets, and prepares all the interaction scripts. The only human step is claiming test ETH from a faucet, which takes two minutes. Everything else is automated.

The current June 2026 pipeline targets ACI (30M token pool), MetaMask Season 2, Hyperliquid Season 2, and several ongoing testnets. Estimated total value across all airdrops: 50 to over 500.

2. Content Publishing

Claude writes and publishes technical articles directly to Dev.to via API. The content covers crypto tutorials, Python guides, and AI tool reviews. Each article builds audience and creates opportunities for sponsorships and freelance leads.

3. Passive Income Infrastructure

I am setting up DePIN networks that pay for idle resources: Grass Network for sharing bandwidth, and exploring GPU compute sharing on Render Network. These run in the background and generate passive crypto income.

The Architecture

Memory System

The biggest challenge with AI agents is memory. Claude Code sessions lose context when they end or when the context window fills up. I solved this with a three-layer memory architecture.

Layer one is a CLAUDE.md file in the project root that Claude reads on every startup. It contains wallet addresses, active airdrop lists, constraints, and key paths. Layer two is MEMORY.md files that persist across sessions. Layer three is an SQLite database with full-text search that auto-syncs conversation history.

Skills That Actually Help

After testing dozens of community skills, I kept fourteen that genuinely improve Claude's performance. The brainstorm and council skills provide structured thinking frameworks for evaluating opportunities. The systematic debugging and TDD skills keep code quality high. The extract-learnings skill automatically distills insights from conversations into persistent memory.

Context Management

Long conversations fill the context window and degrade performance. The solution is simple: save critical information to CLAUDE.md, then use the clear command to reset context. Claude wakes up fresh, reads CLAUDE.md, and continues where you left off.

Results So Far

In the first few days of running this system, I have built a crypto earning dashboard, created dedicated airdrop farming wallets, published technical articles, and deployed a memory system that persists across sessions. Actual airdrop distributions are pending the test ETH faucet step and claim windows opening later this month.

What I Learned

The internet payment infrastructure assumes a human is behind every keyboard. Every platform requires some form of identity verification. This is the real bottleneck for AI agents, not intelligence or capability.

The workaround is to use crypto-native platforms that only require wallet addresses, not identities. DePIN networks, crypto airdrops, and blockchain-based content platforms all fit this model.

Next Steps

I am expanding the pipeline to include more testnets, building GPU compute sharing infrastructure, and exploring AI agent marketplaces where agents can list services and get paid directly in cryptocurrency.

Follow me for updates on this project. I will share real earnings data as airdrop distributions come in.

How I Built a Crypto Airdrop Farming System With Zero Investment (June 2026)

叶稼辰 — Thu, 11 Jun 2026 04:02:58 +0000

How I Built a Crypto Airdrop Farming System With Zero Investment (June 2026)

A step-by-step guide to earning free cryptocurrency through testnet airdrops, written by someone actively farming right now.

Why Airdrops Are Worth Your Time

Crypto projects distribute millions of dollars in free tokens to early adopters and testers. This is not speculation — it is a proven model. Projects like Arbitrum, Optimism, and Aptos have given away thousands of dollars per wallet to early users.

In June 2026, there are multiple active airdrop campaigns with deadlines this month. The best part: you need exactly zero investment to participate.

My Setup

I am running this operation on a standard gaming PC:

Windows 11 with Python 3.11 and web3.py
A fresh Ethereum wallet created specifically for airdrops
No initial capital whatsoever

Active Airdrops in June 2026

ACI Testnet (Deadline: June 30, 2026)

The ACI project has allocated 30 million tokens for Phase 1 testnet participants. To qualify, you need to swap test ETH for ACI tokens, stake them, and complete the available missions. The testnet runs on Ethereum Sepolia, so you only need free test ETH from a faucet.

Estimated value per wallet: $50 to $500 depending on participation level.

MetaMask Season 2 (Claim Window: June 15-30, 2026)

If you have ever used MetaMask to swap tokens or bridge assets, you may be eligible for the Season 2 MASK token claim. Check your eligibility through the MetaMask Portfolio dashboard during the claim window.

Estimated value: $100 to $2000 or more for active users.

Hyperliquid Season 2 (Claim by June 25, 2026)

Hyperliquid is distributing HYPE tokens to traders who have been active on their platform. Even moderate trading activity may qualify. The claim deadline is June 25.

Estimated value: $500 to $5000 depending on trading volume.

Ongoing Testnets

Several projects are running continuous testnet campaigns where you earn points or roles that are expected to convert to tokens at launch. These include Linera, Seismic, and Rome. These take more time but have no deadline pressure.

How to Get Started

1. Create a Fresh Wallet

Always use a dedicated wallet for airdrop farming. Never use your main wallet. Here is the code I use:

from eth_account import Account
import secrets

private_key = "0x" + secrets.token_hex(32)
account = Account.from_key(private_key)
print(f"New wallet address: {account.address}")
# Save the private key securely

2. Get Test ETH

For Ethereum testnets, you need Sepolia ETH which is completely free. Visit the Sepolia PoW Faucet at sepolia-faucet.pk910.de. No registration required, no captcha. Just paste your address and let the browser mine in the background. You can also try faucets from Chainstack, QuickNode, or Alchemy if you prefer.

3. Start Interacting

For each testnet project, the pattern is similar: connect your wallet, perform transactions like swaps and stakes, and complete any available social tasks on platforms like Zealy or Galxe. The more genuine your interactions look, the better your allocation will be.

Pro Tips from My Experience

Use separate wallets for different ecosystems to reduce the risk of being flagged as a Sybil attacker. Spread your interactions across multiple days rather than doing everything at once. Complete every available task and mission. Most importantly, never pay anyone for an airdrop — if someone asks for money, it is a scam.

Expected Earnings

Here is my realistic estimate based on current airdrop values and historical comparisons:

The ACI testnet could yield $50 to $500. MetaMask Season 2 might bring $100 to $2000. Hyperliquid Season 2 could deliver $500 to $5000. The ongoing testnets like Linera and Seismic add another $100 to $1000 in potential value.

Total potential per dedicated farmer: $750 to over $8500, entirely from free airdrops.

Final Thoughts

Airdrop farming is the closest thing to free money in the cryptocurrency world. It requires patience, consistency, and attention to detail, but the upside is real and measurable. With several June deadlines approaching, the best time to start is today.

This article is part of my series on earning cryptocurrency with zero initial investment. I will be sharing updates as I receive airdrop distributions and exploring new opportunities as they emerge.