# chrisworsey55/atlas-gic

ATLAS by General Intelligence Capital — Self-improving AI trading agents using Karpathy-style autoresearch

Repository: https://github.com/chrisworsey55/atlas-gic
Canonical: https://ross.abutalabs.com/products/atlas-gic
Language: Python
License: NOASSERTION
License Family: other
Last push: 2026-05-27T19:40:53+00:00

## Health v2 (maintenance only)
Score: 52/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 84, release rhythm 35, longevity 12
- inputs: {"age_days": 175, "days_push": 98, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, young, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2088, forks 375 (observed 2026-08-28T04:06:12.149734+00:00)

## What it is
ATLAS is a Python framework of self-improving AI trading agents that optimize their own prompts against market outcomes, using multi-agent debate across macro, sector, and market layers. It is tied to a commercial SaaS platform (ATLAS Agents) offering live signals, copy trading via Alpaca and Kalshi, backtesting, and an agent marketplace.

## Use cases
- run self-improving AI trading agents that rewrite their own prompts
- backtest trading strategies described in plain English
- copy-trade AI agent signals into equities and prediction markets
- simulate reflexive market futures before deploying a strategy
- benchmark agent performance using Sharpe ratio as the objective
- publish and monetize custom trading agents on a marketplace

## When to choose
- you want an open framework for autonomous LLM trading agents with prompt evolution
- you need multi-agent market debate with performance-scored feedback loops
- you want to experiment with autoresearch-style optimization applied to finance

## When to avoid
- you need a battle-tested, audited trading system for serious capital
- you want a neutral open-source tool without commercial upsell pressure
- you require guaranteed returns or verified track records from marketing claims

## Facets
- artifact type: framework
- maturity: active
- function: agent-framework, machine-learning, trading, benchmarking, workflow-automation
- domain: fintech, artificial-intelligence, large-language-models
- platform: python, cli, cross-platform
- tags: ai-trading-agents, autoresearch, prompt-optimization, self-improving-agents, backtesting, copy-trading, prediction-markets, reflexivity-simulation, multi-agent-debate, saas, ai-agents, trading

## Member repositories
- chrisworsey55/atlas-gic (main) score 52

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:12.149734+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:55:41.375843+00:00, confidence not recorded.
  - readme: https://github.com/chrisworsey55/atlas-gic (fetched 2026-08-28T04:06:12.149734+00:00, sha 61660fefc066)
- Data as of 2026-08-30T08:39:29.467469+00:00.