AutoPredict

Read-only market scanning and provenance-aware forecast evaluation framework

Website checked 6 Nov 2026Free optionPython

AutoPredict supplies public-market scanning, typed snapshots and forecast evaluation workflows. Current defaults use market-implied no-edge forecasts until verified data is supplied. The release explicitly has no authenticated trading extra, and its audit rejects every live configuration even when credentials are present.[1]

Written by Vultax from the product's own pages, last checked .

Category
AI agents
Pricing
Free option
Works with
Polymarket
Runs on
Python
First recorded
27 Aug 2026
Last checked
6 Nov 2026
Saved profiles

Details

Type
Software repository

Timeline

  1. 6 November 2026

    Website answered

  2. 30 October 2026

    Website first checked by Vultax

  3. 27 August 2026

    Listed in the Kairos catalogue

Quick answers

What does it cost?
The reviewed repository documentation does not publish a complete software price or current venue-fee schedule; any venue and deployment costs remain separate.[1]
Does it trade for me?
Read-only live scanning and offline evaluation; no live execution in current release.[1]
Where do my funds sit?
No current trading integration; credentials do not enable execution in this release.[1]
What do I need to start?
Python environment; user-supplied real historical observation/resolution data for evaluation; no authentication needed for public scans.[1]
Where is it available?
The repository documents a self-operated software workflow, not a complete country-eligibility list. Underlying venue and service restrictions still apply.[1]

How it works

Trading mode
Mode:
Read-only live scanning and offline evaluation; no live execution in current release.
Delivery channels
CLI, Python framework, Evaluation reports

Getting started

  1. Install the Python package from the repository.

  2. Use public live scanning or supply a versioned real-observation dataset manifest.

  3. Validate the dataset and evaluate against a baseline; inspect provenance and warnings.

What you need first
Requirements:
Python environment; user-supplied real historical observation/resolution data for evaluation; no authentication needed for public scans.

Safety and terms

Custody and keys
No current trading integration; credentials do not enable execution in this release.[1]
Terms, privacy and eligibility

Sources

  1. 1. github.com/howdymary/autopredict (opens in a new tab)

Download profile JSON. Vultax keeps a dated copy of every page behind it. Request a review.

Listings that share its category or the jobs it does.

Earlier descriptions come from the Kairos catalogue by its contributors, used under CC BY 4.0. Attribution and licences.

How the directory is made · Manage a profile · Directory CSV (CC BY 4.0).