skills.rest
surfingalien
surfingalienCOMMUNITY
Explore all skills by this author →
GitHub

autopredict

Scan Polymarket markets, backtest trading strategies, and tune execution parameters via helper scripts.

1|Updated May 6, 2026
Search Tags:#bash#backtesting#trading strategy#polymarket#parameter tuning#prediction markets#market scanning
One-click install
npx skills add https://github.com/surfingalien/FinSurfing --skill autopredict-surfingalien
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: autopredict
Source: https://github.com/surfingalien/FinSurfing/tree/main/.claude/skills/autopredict
Command: npx skills add https://github.com/surfingalien/FinSurfing --skill autopredict-surfingalien

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve? The upstream howdymary/autopredict repository has broken defaults, stub CLI commands, and scaffold-only live trading code, making it easy to run invalid backtests or misjudge what the repo can actually do. This Skill wraps the repo with audited guidance and safer helper scripts so you can scan markets, backtest, and tune parameters without hitting known failure modes. ## Core Features & Use Cases - Live Market Scanning: Scan live Polymarket markets, inspect multi-outcome event overround/underround, and evaluate a single market against your own fair probability estimate. - Reproducible Backtesting: Run backtests against JSON datasets with guaranteed dataset handling, avoiding the upstream TypeError caused by a null default dataset. - Parameter Tuning: Grid-search strategy parameters like min_edge and aggressive_edge with ranked metrics output, bypassing the stub learn tune CLI. - Use Case: You have a hypothesis that lowering min_edge improves Sharpe on a market snapshot dataset. Use the tuning helper to sweep min_edge values, then compare summary.json results across runs. ## Quick Start Ask the assistant to set up the AutoPredict repo and scan the top 10 most liquid Polymarket markets with verbose execution details.

Frequently Asked Questions about autopredict

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I scan live Polymarket markets from the command line?▼

Run predict.py with flags like --top 10 --verbose to list the most active markets with spreads, depth, and overround signals. Live reads require internet access but no credentials. The scan_markets.sh helper wraps these calls with repo validation.

How do I backtest a Polymarket trading strategy config?▼

Run python3 -m autopredict.cli backtest with both --config and --dataset arguments. Always pass a dataset explicitly because the repo's default_dataset is null and omitting it throws a TypeError. The bundled sample_markets.json works as a smoke-test input.

Why does the AutoPredict backtest fail with a TypeError before starting?▼

The TypeError occurs because no --dataset was passed and the upstream config.json sets default_dataset to null, which the code tries to resolve as a path. Fix it by passing a valid dataset JSON file to the backtest command.

Can AutoPredict place real trades on Polymarket?▼

No, live trading is scaffold-only in the current upstream repo. The run_live.py script uses a MockVenueAdapter, the trade-live CLI is disabled by config, and learn tune/improve are placeholders pointing to a nonexistent script.

Does predict.py --fair use my strategy config file?▼

No, predict.py --fair constructs the agent with the default AgentConfig and ignores strategy_configs/*.json files. Treat its output as a default-policy sanity check rather than validation of a tuned configuration.

How do I tune AutoPredict strategy parameters like min_edge?▼

Use the tune_params.sh helper with repeatable --param flags specifying comma-separated values, plus a dataset and scoring metric such as sharpe. It grid-searches combinations via the working backtest CLI and writes ranked results to summary.json.

Related Skills

Explore More Finance & Accounting Skills→
serenorgserenorg
6

Polymarket Bot

Automate Polymarket prediction market trades using AI research and Kelly Criterion.

Official
Advanced
adennngadennng
33

multi-attempt-comparison

Compare and rank strategy attempts by Sharpe ratio, total return, and drawdown.

Community
Advanced
thistleknotthistleknot

mlflow

Persist MLflow experiment parameters, metrics, artifacts, and nested run lineage.

Community
Intermediate
GhostOf0daysGhostOf0days

algorithmic-trading-dma-handbook

Automates algorithmic trading DMA handbook workflows for production systems.

Community
Advanced