Research-grade synthetic data generator for finance with 18 dataset profiles, Gaussian copula + VECM/GARCH models, fidelity + privacy audit framework, typed Python, comprehensive docs (design.md, ARCHITECTURE.md, STATUS.md), 168 passing tests, and Flask REST API. Early-stage active portfolio project under JOSS review.

ML-powered financial modeling toolkit (DCF, GARCH, options, credit, risk) with 58 modules across 13 layers. Typed Python, CI/CD, tests, MIT license, ~400KB codebase. Early-stage indie project (5 stars, 40 days old) shipping structured, documented modules but pre-market adoption.

Novel MARL framework for adversarial market microstructure with MINE-based information leakage penalty, featuring realistic LOB matching, belief inference, and alternating optimization. Early-stage research codebase (1 star, created March 2026) with solid engineering (typed Python, CI/tests, docs) but minimal external

A rigorous formal-theory implementation of CHSE (Contested Hierarchy with Social Embedding) game theory in Python. Well-documented with multi-year scope, typed code, but nascent adoption (0 stars, 9 days old, sparse test coverage of advanced features).

From-scratch GBDT library with finance-specific losses (Sharpe, PinballLoss, DrawdownLoss), Mondrian conformal prediction, and Regime-Weighted Gradient Boosting. Novel algorithms validated on regime-dependent benchmarks; well-documented with tests, CI, and organized multi-file architecture (158 KB, HAS_CI=yes, HAS_TEST

Early-stage Python mutual fund analytics library with typed code, structured multi-file layout, tests, CI, and detailed README. Created 2026-05-02, minimal commit history (5 of last 30), but ships with complete API surface: NAV fetching, analytics (CAGR, drawdown, rolling metrics), CLI, and comprehensive test suite.

Brand-new financial benchmarking library with regime-aware performance metrics, HMM-based regime detection, and multi-strategy comparison. Well-structured codebase with comprehensive tests and CI, but zero adoption signals (no stars, no external visibility yet).

Early-stage ML backtesting library (v0.6.0) for single-stock analysis with 44 features, 14 models, and AutoPilot automation. Brand new repo (4 hours old) with minimal adoption but solid technical foundation: typed Python, structured src/, comprehensive docs, tests, CI/CD. Experimental prototype phase.

Experimental finance ML library (FORGE) with sophisticated architecture but zero adoption (0 stars, <24h old, 1 commit). Typed Python + docs + tests + CI present; however incomplete implementation, no real validation data, and no production evidence.

Early-stage Python research framework for alpha modeling with ambitious design docs but minimal implementation (3 commits, 80 KB total, untyped). Shows promise via comprehensive README and architecture docs but lacks substance: no source files sampled, 0 stars/forks, created today.