Skip to content

Latest commit

 

History

History
29 lines (22 loc) · 2 KB

File metadata and controls

29 lines (22 loc) · 2 KB

Description

This PR delivers a comprehensive README overhaul to reflect the Phase 7 Machine Learning Integration. It transforms the documentation from a pre-ML state to a world-class README that accurately represents AltFlex as a full-stack exploit intelligence platform with ML-powered pattern recognition.

Changes

  • Badge Row: Added XGBoost, scikit-learn, ONNX Runtime, and Python badges.
  • Overview Section: Added the "🧠 ML Exploit Pattern Recognizer" module and expanded the Module Capability Matrix to 3 columns.
  • Key Features: Added "🧠 Machine Learning Intelligence" block with 6 feature highlights.
  • Architecture Diagram: Updated the Mermaid graph with a new ML Intelligence subgraph (OnnxExploitClassifier, TraceFeatureExtractor), ONNX model node, and connection arrows.
  • Tech Stack: Added 6 ML-specific rows (XGBoost, scikit-learn, ONNX Runtime, pandas/numpy, matplotlib/seaborn, Python).
  • Monorepo Structure: Expanded the tree to show adapters/ml/, the full research/ sub-structure (datasets, models, figures, reports), and scripts/ml/.
  • ML Model Performance (NEW section): Embedded the Heuristic vs. XGBoost comparison table, per-pattern Δ metrics, and all 6 thesis figures in a 2×3 grid.
  • Phase Roadmap: Added Phase 7 row and full Phase 7 Task Tracker (7 tasks with PR links).
  • Academic Alignment: Added Phase 7 → Thesis 1 & 2 mapping.
  • Package Dependency Graph: Updated @aegis/forensic-engine to show onnxruntime-node dependency and scripts/ml/ → ONNX model pipeline.
  • Changelogs: Added [03.7.0] entry documenting all ML sub-tasks (P7-ML-001 through P7-ML-007).
  • Table of Contents: Added "ML Model Performance" entry.

Type of change

  • Documentation update (non-breaking, no code changes)

Checklist:

  • My changes follow the project's documentation style
  • All internal links resolve correctly
  • Mermaid diagrams render without errors
  • Embedded figures reference existing files in research/figures/