Milli.aiFree guides

The POLYBOT 6-Agent Blueprint

A Polymarket bot reportedly turned $150 into $279,053. I rebuilt it for you. Here's how its 6 AI agents work, how to build it, and a free prompt that runs the whole team in 10 minutes.

How it works: Scout finds markets → Odds finds the gap → News checks the facts → Thesis builds both sides → Risk can cancel → you approve

READ FIRST: It never places a bet on its own. Most ideas get cancelled. Check that prediction markets are legal where you live.

Free full guide (step by step, with 12 copy-paste prompts, daily routine and weekly review): milli-ai.com/free/polybot

Get the full free guide →

How it works

  1. SCOUT checks the 100 busiest markets on Polymarket (free public data). It flags 3 things: a big jump in betting, odds moving 5+ points, or an end date coming soon.
  2. NEWS searches the web for fresh reporting. It only counts facts with a source and a date.
  3. THESIS builds the best case for YES and for NO, and names what would kill the bet.
  4. ODDS is simple math. If the market says 42% and the team says 57%, that's a 15-point gap.
  5. RISK is the bouncer. It sizes the bet small (never more than 2% of your money) and blocks anything that fails a check: gap too small, news is a rumor, not enough money in the market, or too little time left.
  6. EXECUTION writes one ticket and sends it to your phone. It never bets for you.

Every run is saved to a log. After a month, you can see if the team beat the market. If it didn't, you don't trust it. Simple.

Educational only. Not financial advice. You can lose what you bet. Check that Polymarket is allowed where you live.

Build the free starter version with Claude Code (20 minutes)

Want it running on your computer instead of in a chat? Paste one prompt and Claude Code builds a starter POLYBOT that scans Polymarket markets and runs the 6 agents on the best ones. It runs on pretend money or in read-only mode. It never places a trade or bet.

You need: Claude Code (needs a paid Claude plan), Python 3, and an Anthropic API key from console.anthropic.com (pay as you go, usually cents per run).

1. Install Claude Code (Mac/Linux terminal):

curl -fsSL https://claude.ai/install.sh | bash

Windows PowerShell: irm https://claude.ai/install.ps1 | iex

2. Make a folder and open Claude Code in it:

mkdir polybot && cd polybot
claude

3. Paste this build prompt:

Build me a beginner-friendly starter bot called polybot in Python 3: a starter POLYBOT that scans Polymarket markets and runs the 6 agents on the best ones.

WHAT IT DOES
1. Pull active markets from the public Polymarket Gamma API (https://gamma-api.polymarket.com/markets?active=true&closed=false). If a field name is unclear, print a sample and leave a clear TODO instead of guessing.
2. Keep markets with enough volume that resolve within 60 days. For the top 5, run 6 Claude roles: Scout, Odds, News, Thesis, Risk (can veto), Execution (writes the ticket only). Each replies as JSON.
3. Never place bets. No wallet, no keys, no signing code. Log every ticket to logs/tickets.jsonl.

RULES FOR THE BUILD
4. Use the official Anthropic Python SDK. Put the model name in config.json. Read ANTHROPIC_API_KEY from .env with python-dotenv. Create .env.example with placeholders and add .env, data/ and logs/ to .gitignore. Never hardcode keys.
5. It never places real trades or bets. No broker or wallet code, no private keys.
6. Put every setting (limits, tickers, thresholds) in config.json so I can change them without touching code.
7. Add a --demo flag that runs the whole thing on made-up data with no internet, so I can test it first.
8. Handle errors and rate limits with a short wait and retry. If you are not sure an API endpoint or field exists, leave a clear TODO instead of guessing.
9. Write a README.md for a total beginner: install, add the key, run --demo, run for real, read the logs.
10. Keep it simple and well commented, under about 400 lines. When done, explain what you built and list anything you could not check.

4. Install and test:

pip install -r requirements.txt
cp .env.example .env
python src/main.py --demo

Add your key to .env with a text editor, then run it for real (the README shows the exact command).

Keep your key safe: never paste API keys into a chat, a screenshot or GitHub. Keep .env in .gitignore.

Starter vs finished

Starter (free, this prompt) Finished (AI Desk)
Works on your computer ✅ ✅
Runs every hour and texts you the best tickets ❌ ✅
News checks with dated sources ❌ ✅
Self-grading log so you see if it works ❌ ✅
Tested agent prompts ❌ ✅
Tested by us, with setup support ❌ ✅

Troubleshooting

Problem Fix
It treats rumors as news Paste only dated, linked sources.
Every market looks good Ask Risk: "Give 3 reasons to skip this."
It can't see live prices Paste the current price and volume yourself.

Want the finished version?

The prompt above builds a starter. The finished, tested version (one-command install, plain setup steps, and help if you get stuck) is part of the Milli.ai AI Desk.

Educational only. Not financial advice. It never places trades for you. You make every decision.