Sessionized identity evidence
Converts CloudTrail, Okta, Entra, GitHub, and local provider exports into a stable session contract for scoring and analyst review.
The Generative Identity Forensics and Trust System (GIFTS) turns provider-style audit events into sessionized identity findings, analyst-readable reports, local dashboards, and dry-run response plans without executing provider actions.
GIFTS is currently a reference implementation for cloud identity evidence review. It sessionizes audit events, scores suspicious identity behavior, records what is known and uncertain, and keeps response activity in a governed dry-run lane.
Converts CloudTrail, Okta, Entra, GitHub, and local provider exports into a stable session contract for scoring and analyst review.
Produces JSONL, CSV, markdown reports, dashboards, and triage snapshots that explain why a session needs attention.
Builds response previews and execution ledgers for manual or test-tenant review while keeping provider actions disabled in the public reference package.
Tracks public datasets, disposable lab traces, conversion-loss reports, and source manifests so future benchmark claims stay reviewable.
The paper motivates geodesic anomaly detection and manifold diffusion. The current implementation keeps that path demo-scale and labels reconstruction output as analyst-support hypotheses, not facts.
The May 2026 implementation adds a product-shaped cloud identity incident demo, public/lab source cataloging, Splunk public-data validation, and conversion-cap audit reporting. These are still reference-implementation results, not production accuracy claims.
This public release stays focused on the materials most useful to first-time visitors: the current prototype, the implementation evidence, the publications, and the contact page.
Use the publications, prototype, updates, and contact pages as the main public entry points into the current GIFTS materials.
GIFTS is public technical work by Kingdom Mutala Akugri, whose focus areas include cloud identity and security engineering, anomaly-detection methods, and applied security evaluation. This site is intended to provide a clear public entry point for the research lineage, prototype materials, and technical framing behind GIFTS.
Practical identity assurance depends on measurable controls, repeatable evaluation, and deployment realism. GIFTS is designed around that engineering mindset rather than around abstract policy statements alone.
The framework is intended to produce reusable logic, reference architectures, and evaluation guidance that can be studied and adapted across organizations without requiring one employer-specific implementation.