AEGIS Core — Governance Enforcement Engine (Python Reference Implementation)

Security-hardened Python reference implementation of the AEGIS Governance Enforcement Engine. Implements the AEGIS Governance Protocol (AGP-1) message processing pipeline with a 3-stage governance evaluation: capability check, policy evaluation, and risk scoring. Version 0.1.1b2 includes 353 tests across 9 rounds of adversarial red/blue team testing grounded in the AEGIS Threat Model (ATM-1) and AEGIS Threat Matrix (ATX-1 v2.1). All 29 applicable ATX-1 techniques are covered at the engine layer. 30+ security fixes applied across two independent adversarial sessions. The risk engine provides 5-dimension composite scoring (0.0–10.0, AGP-1 aligned) with a proportionality gate that escalates destructive actions for human review. Scoring dimensions: capability sensitivity, action severity, target sensitivity, historical attempt rate, and behavioral anomaly detection. This artifact serves as Tier 1 primary empirical evidence for the ATX-1 taxonomy, validating that adversarial testing of the reference implementation produced zero taxonomy gaps across all applicable techniques.