
Welcome to the course! Meet your instructor and learn what you'll build: an AI crypto trading screener that finds coins, sets entry and exit levels, and sends daily trade picks to Telegram, all with Claude Code and no coding. See why this course is short and step by step instead of a 12-hour marathon, and how to get the most out of it: download the prompt templates and starter files, follow along step bu step.
Download and install the Claude desktop app from claude.com/download, sign in with your paid Claude account (Pro or Max) and open the Code tab: Claude Code, your AI trading assistant for the rest of the course. A quick tour of sessions, the chat box, permission prompts and the file view, plus what curl and an API are. Resource: step-by-step install guide with links.
Create a project folder, trust the workspace, then one plain-English prompt creates your whole project: CLAUDE.md titled House Rules, checklist.md titled Signal Checklist, a .claude/commands folder for your screeners, a results folder for daily signals and a private folder for secrets. Learn what each part does, plus your first new term: a .md file is a Markdown file, plain text with simple formatting like # titles, which Claude reads as instructions. The two files start empty and fill up as you build. Resource: the project prompt to copy and paste.
With one plain-English prompt, Claude fills your project: CLAUDE.md house rules (KuCoin USDT pairs, curl only, a $1,000 account and 1% risk) and checklist.md with the 7-point trade checklist. Learn what each rule does and why the rules file keeps every answer consistent. Resource: the rules prompt to copy and paste.
Start with a quick recap: your .md files are plain-text Markdown that Claude reads as instructions, so CLAUDE.md holds your house rules and checklist.md your 7 checks. Then one new term: an MCP (Model Context Protocol) server is a connector that lets Claude use another app, here your TradingView chart. Create a free TradingView account, install TradingView Desktop and open a KUCOIN:BTCUSDT chart (free plan: 2 indicators per chart, 3 alerts, unlimited drawings). Then install Node.js (LTS) and give Claude one prompt that downloads the free tradesdontlie TradingView MCP from GitHub, installs it and connects it to Claude Code through a local debug port. No coding: Claude runs every step and you approve. Resource: install prompt, links and fixes for common errors.
Ask Claude to pull live prices for every KuCoin USDT pair with one curl call, skip stablecoins and leveraged tokens, find the top gainer with over $10M of 24h volume and switch TradingView to that coin on the 4-hour chart. Learn how to read the reply and check the numbers yourself. Resource: the prompt and the KuCoin endpoint.
Five questions on the Claude Code tab, what CLAUDE.md and checklist.md do, the 1% risk rule, what an MCP server is and how to check the TradingView connection.
Define each signal with clear numbers: momentum is up 3–15% today on over $10M volume, a breakout trades within 1% of its 24h high, a dip is down 5–15% while the daily trend still holds, and a range has a 24h high-to-low gap under 4% (ideal for grid bots). Learn when each one works and when it traps traders. Resource: the 4-signal cheat sheet.
Have the AI add Volume and a 20 EMA (within the free plan's 2-indicator limit) and judge whether a signal is confirmed: price above the 20-day average and volume above its 20-day average. Then apply the three filters that remove most failed signals: liquidity (over $10M 24h volume, no stablecoins or leveraged tokens), not chasing (up less than 15% today) and market mood (skip longs when BTC is down more than 3%).
Pull free market context with curl, no API key: the Crypto Fear & Greed Index (alternative.me), CoinGecko trending coins and BTC dominance, headlines from the Cointelegraph and Decrypt news feeds, and KuCoin new-listing and delisting announcements. Learn to read sentiment extremes, treat a fresh catalyst as a bonus and never a reason to chase, and auto-skip any coin with a hack, exploit, delisting or token-unlock headline. Resource: the news and sentiment source sheet with every endpoint.
Six questions on signal types, volume and trend confirmation, the three failure filters and reading news and sentiment.
Build a screener prompt in four parts: the data source (KuCoin's allTickers endpoint returns every pair in one call), the signal rules, the filters with the news and sentiment check, and the exact output table. Run the momentum scan, pull 20 daily candles for each finalist and let the AI run the 7-point checklist to mark every coin TAKE, WAIT or SKIP. Then save it in .claude/commands so /scan, /scan-breakout, /scan-dips and /scan-sideways each run a full scan. Resource: all four command files.
Five questions on the four parts of a screener prompt, the allTickers endpoint, TAKE, WAIT and SKIP verdicts and saving scans in .claude/commands.
Create a weighted scoring prompt that builds on the checklist: trend 20, volume 20, momentum 15, market regime 15 (BTC trend plus Fear & Greed), news and trend catalyst 15 and reward-to-risk 15. A risky headline (hack, exploit, delisting, unlock) caps the grade at C. The AI grades every result A (75+), B (60–74) or C (below 60) and explains each score. Resource: the AI Signal Score sheet.
Have the AI add every A-grade coin to your TradingView watchlist as KUCOIN USDT symbols in score order. Then it opens each coin, flips through the daily, 4-hour and 1-hour charts, checks the signal bar and the two indicators, and confirms or downgrades the setup when the higher timeframe disagrees.
Five questions on the score weights, the A, B and C cut-offs, the risky-headline cap and when a higher timeframe downgrades a setup.
Use the formula position size = dollar risk ÷ (entry − stop) and have the AI calculate it for every pick, with a $250 cap per coin and a stop no more than 5% from entry. Example: $10 risk, entry $2.00, stop $1.90 = 100 coins, a $200 position. Resource: position-size cheat sheet.
The AI places the stop below the last swing low or 1.5× ATR(14), whichever is tighter within 5%, sets targets at 2R and 3R and draws colored lines and a label with the position size on the TradingView chart. Drawing tools are unlimited on the free plan.
Five calculation questions on position size, stop distance and reward-to-risk ratios.
You need a free Telegram account (phone app plus Telegram Desktop). Step by step: message @BotFather, send /newbot, choose a name and a username ending in "bot" and copy the token. Send your bot a message so Claude can find your chat ID with one curl call, then save the token and chat ID in private/telegram.md (never share it) and add the Telegram rule to CLAUDE.md. Claude sends a test message, then captures the drawn TradingView chart and sends it with the score, headline, entry, stop, targets and size. Telegram replaces the free plan's 3-alert limit. Resource: bot setup guide and test commands.
Save each scan to results/YYYY-MM-DD.md, then have the AI reopen past signals in TradingView, check whether the target or stop was hit first and report win rate and average R by grade, by market mood and with or without a news catalyst, so you keep only what works. Resource: daily journal template.
Five questions on BotFather setup, keeping the token and chat ID private, what a chart alert contains and reading win rate and average R by grade.
One prompt checks the market mood and news, scans every USDT pair, scores the signals, fills the TradingView watchlist, tours each chart, draws the levels with 1% sizing, sends the charts to Telegram and saves the day's results. Then compare it with the same work done by hand. Resource: the one-prompt workflow and the full prompt pack.
Five questions on the order of the full workflow, the free-plan limits it works around and which decisions always stay with you.
Build five extra screeners as one-word commands: Trend Pullback, Oversold Bounce, Squeeze Breakout, Volume Surge and Strength vs BTC. Learn the rule behind each one, when it works, and how each fills your TradingView watchlist with entry, stop and target levels. Resource: all five command files.
Recap the automated workflow and the five screener strategies you built, see ideas for new scans, news sources and chart automations, and learn where to get help in the Q&A.
This course contains the use of artificial intelligence.
Bull market or bear market, what if one AI prompt could scan the entire cryptocurrency market and hand you the best trade setups in minutes?
Welcome to AI Crypto Trading for Beginners, a step-by-step introduction to using artificial intelligence, Claude and TradingView to find, analyze and plan cryptocurrency trades with automation.
Most beginners trade crypto by guesswork. They type tickers one by one, flip through hundreds of charts, buy coins that already pumped, miss the news behind a move and have no plan for when to get out. This course replaces that guesswork with a clear, rules-based crypto trading strategy that AI runs for you.
You'll build your own AI crypto screener. It scans every KuCoin USDT pair, checks the market trend and the news, scores every trading signal, fills your TradingView watchlist, draws your trade levels on the chart and sends the best setups to your phone through Telegram alerts. You do it all with plain-English prompts. You write zero code.
What you'll learn, section by section
➤ Section 1, Welcome and Setup: install the Claude desktop app and open Claude Code, build your screener project with one prompt, write your house rules and 7-point checklist, install TradingView Desktop and the TradingView MCP, then type one prompt and watch AI take over your TradingView chart
➤ Section 2, Screener Signal Foundations: learn the 4 screener signals that matter (momentum, breakout, dip and range), confirm them with volume and the 20 EMA trend, filter out fakes with 3 filters and add news, trending coins and market mood to every crypto signal
➤ Section 3, Build Your AI Crypto Screener: build one market scan prompt that covers the whole crypto market and save it as one-word screener commands for momentum, breakout, dip and sideways markets
➤ Section 4, AI Signal Score and Watchlists: build an AI Signal Score that grades every pick from 0 to 100, auto-fill your TradingView watchlist and let AI tour every A-grade chart on the daily, 4-hour and 1-hour timeframes
➤ Section 5, Risk Management: size every position with the 1% rule on a $1,000 account and let AI draw entry, stop-loss and take-profit levels on your chart
➤ Section 6, Telegram Signal Alerts: send A-grade signals with chart screenshots to Telegram and save a daily signal journal that tracks win rate by grade, so you keep the strategies that prove profitable and drop the ones that don't
➤ Section 7, Final Project: one prompt, full autopilot. Scan, score, chart, draw and alert in one run. Then get 5 crypto screener strategies that auto-fill your TradingView watchlist with entry, stop-loss and take-profit levels.
Why this course is different
➤ Short and practical: about 1.5 hours, not a 12-hour marathon
➤ Built for beginners: no coding, no Python and no trading experience needed
➤ Copy-and-paste prompts: lectures come with the exact prompt plus a one-page resource guide
➤ Free tools and data: a free TradingView plan, free Telegram and free public crypto market data
➤ A quiz at the end of every section: 7 quizzes to check what you learned before moving on
Who this course is for
➤ Beginners who want an easy introduction to cryptocurrency trading with AI
➤ Crypto traders tired of doing chart analysis and market screening by hand
➤ Swing traders who want a faster way to find trending coins and plan positions
➤ Anyone curious about AI, Claude, prompt engineering and trading automation
➤ People who want a disciplined trading process for bull and bear markets before they think about profitable trading or passive income
What you need
A paid Claude plan (Pro or Max), a free TradingView account, a free Telegram account and a computer. That's it.
An honest note: this course teaches tools, analysis and process, not a get-rich-quick system. No screener or AI makes trading profitable on its own, and the cryptocurrency market is risky. You'll practice with a $1,000 account and 1% risk per trade, and you always make the final decision.
Enroll now and let AI scan the crypto market for you, so you can focus on the decisions that matter.