Beginner 10-15 minutes

Your First CLI Trade: Safe Beginner Runbook (10-15 Minutes)

Updated

TL;DR: If you are new to terminal trading, do this in order: install one CLI, set strict risk limits, run a market scan with your AI agent, place one tiny test order, and log the trade. This guide gives you a repeatable first-trade workflow that is actually safe and useful, not just a demo.

Who This Guide Is For

This guide is for:

  • First-time CLI traders who want a real workflow, not a gimmick
  • Non-programmers who can follow simple command and prompt examples
  • Traders who want agent-assisted execution with tight guardrails

If you want broader context first, read What Is CLI Trading?, then come back here.

What You Need Before You Start

  • A Mac or Linux machine (Windows with WSL is also fine)
  • Claude Code installed
  • A small, predefined risk budget for this test (recommended: $10-$50 max)
  • 10-15 focused minutes

Safety First: 6 Non-Negotiable Rules

Before placing any order, commit to these rules:

  1. Use only a small test amount for your first session
  2. Do not use leverage
  3. Use markets with clear resolution criteria
  4. Never place a trade you cannot explain in one sentence
  5. Always do a dry-run prompt before an execute prompt
  6. Log every trade so you can learn from outcomes

These rules protect you from the two biggest beginner mistakes: oversizing and impulsive execution.

Step 1: Install and Verify the CLI

Open your terminal and install the tool:

brew install polymarket

Verify it:

polymarket --version

If you see a version number, you are ready. If not, check your shell PATH and rerun the install command.

Related tool page: Polymarket CLI

Step 2: Open Claude Code and Set Guardrails

Launch Claude Code:

claude

Then send this guardrail prompt first:

You are my trading assistant.
Rules:
1) Use read-only analysis first.
2) Do not place any order unless I explicitly say "EXECUTE".
3) Keep first trade size <= $25.
4) Show me market, spread, volume, and resolution criteria before suggesting any trade.
5) Explain downside in plain English.

This single prompt dramatically reduces accidental orders and low-quality analysis.

Step 3: Scan Markets with a Structured Prompt

Now ask Claude:

Show the top 10 most active Polymarket markets.
For each market, include:
- current YES and NO prices
- 24h volume
- bid/ask spread (if available)
- one-line explanation of what resolves the market
Then rank the top 3 by clarity and liquidity for a beginner.

You are looking for markets with:

  • Clear resolution wording
  • Enough liquidity to avoid large slippage
  • Reasonable spread
  • A time horizon you understand

If the output is messy, ask Claude to present it in a simple table.

Step 4: Evaluate One Market Before Trading

Pick one market and run this analysis prompt:

Analyze this market as if I am a cautious beginner:
1) What is the exact claim being priced?
2) What evidence would change conviction?
3) What is the strongest argument against the trade?
4) What are 3 failure scenarios?
5) Give a position size recommendation for a $25 max-risk budget.

Use this quick quality checklist:

CheckWhy It MattersBeginner Threshold
Resolution clarityAvoid ambiguous outcomesWording is explicit and unambiguous
LiquidityReduces slippagePrefer higher-volume markets
SpreadBetter entry pricingPrefer tighter spreads
Time horizonEasier risk controlShort enough to monitor comfortably
Thesis clarityPrevents random betsCan explain edge in one sentence

If any row fails, skip the market.

Step 5: Dry-Run the Order, Then Execute a Tiny Trade

First, ask for a dry run:

Prepare the exact command for a $10-$25 test order.
Do not execute it. I want to review first.
Include:
- side
- amount
- current price
- worst-case loss
- what could invalidate the thesis

Once reviewed, execute:

EXECUTE the previously proposed test order exactly as written.
Then confirm execution details and summarize post-trade risk.

The goal of trade #1 is process validation, not profit maximization.

Step 6: Journal the Trade (Critical for Improvement)

Immediately after execution, store a short post-trade log:

# Trade Journal Entry
- Date/Time:
- Market:
- Thesis (1 sentence):
- Entry price:
- Size:
- Max loss:
- Invalidation condition:
- What would I do differently next time?

Without journaling, you cannot separate luck from skill over time.

Common Problems and Fast Fixes

ProblemLikely CauseFast Fix
command not found: polymarketCLI not in PATHRestart shell and rerun install
Agent gives vague analysisPrompt too open-endedAsk for explicit fields and table output
Fear right before entrySize is too largeCut size until emotional pressure drops
Cannot explain your thesisYou are copying momentumSkip the trade and reassess
Regret after orderNo invalidation ruleDefine exit conditions before next trade

A 7-Day Beginner Progression

If you want fast improvement, follow this sequence:

  1. Day 1: One tiny test trade with full journal entry
  2. Day 2: No trade, only analysis of 5 markets
  3. Day 3: One trade with pre-defined invalidation
  4. Day 4: Review both trades and extract 3 lessons
  5. Day 5: Repeat with stricter sizing discipline
  6. Day 6: Compare your notes to market outcomes
  7. Day 7: Write a one-page personal trading checklist

This builds process quality before increasing risk.

FAQ

Do I need to know programming?

No. You need to run a few commands and write clear prompts. That is enough to start.

Should I start with stocks or prediction markets?

Prediction markets are often simpler for a first workflow. If you prefer equities/options, start with Alpaca MCP Server and use the same guardrail pattern.

What is the biggest beginner mistake in CLI trading?

Treating AI speed as trading edge. Your edge is process quality: clear thesis, small size, repeatable risk control.

What to Do Next

A Note on Risk

This guide uses prediction markets because they are accessible and easy to understand. The risk is still real.

Start small, stay process-focused, and increase size only after your journal proves consistent decision quality. See our full risk disclaimer.