howdymary

howdymary / autopredict

Public

A framework for self-improving prediction market trading agents.

44
7
100% credibility
Found Mar 28, 2026 at 44 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

AutoPredict is an open-source framework for scanning prediction markets, evaluating trade opportunities based on user-provided probability estimates, and simulating or executing bets with execution-aware metrics.

How It Works

1
👀 Discover AutoPredict

You hear about this helpful tool that scans prediction markets and gives smart betting advice based on your own hunches.

2
📱 Set it up easily

Follow a few simple steps on your computer to get everything ready to explore live markets.

3
🔍 Scan live markets

See current betting opportunities, prices, and liquidity from popular prediction sites like Polymarket.

4
🎯 Share your prediction

Enter what you believe the true chance is for an event, like an election or sports outcome.

5
💡 Get personalized advice

It analyzes the market depth and costs to suggest the best bet size, timing, and approach.

6
🧪 Test with simulations

Practice with fake money on past or live data to see how your predictions perform.

📈 Track your smart wins

Watch your betting results improve as you refine predictions and follow the safe advice.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 44 to 44 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 autopredict?

Autopredict is a Python framework for building self-improving trading agents on prediction markets like Polymarket. You supply probability forecasts via CLI or config, and it scans live order books, spots edges like sibling mispricings in multi-outcome events, and proposes execution-aware trades factoring in spread, liquidity, and slippage. Backtest against historical snapshots with realistic metrics for PnL, Sharpe, Brier scores, and fill rates—no prediction generation, just smart execution.

Why is it gaining traction?

In a sea of generic trading libs, autopredict stands out with Polymarket-native API hooks for real-time Gamma/CLOB data and declarative agent configs you tune via JSON for min-edge thresholds or risk fractions. The self-improving loop—backtest, analyze weaknesses like high slippage, auto-suggest tweaks—feels DSPy-inspired for agents, letting you iterate fast without rewriting core logic. Python simplicity plus MIT license makes it a quick fork for custom strategies.

Who should use this?

Quant traders automating Polymarket positions based on personal models, prediction market researchers backtesting execution algos, or agent builders experimenting with self-improving trading loops. Ideal if you're feeding fair probs from LLMs or spreadsheets and need liquidity-aware sizing over raw signals.

Verdict

Grab it if prediction markets are your playground—solid CLI for live scans and backtests punches above its 44 stars—but 1.0% credibility flags early maturity with thin tests. Fork and contribute; docs guide quick wins, but expect to harden for prod.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.