Operational overview

Sovereign runtime: 2M input context on Skia-Serve models (skia:70b, skia:72b, skia:sovereign). Long-context axis routes to skia:llama4-latest at 10M.

All systems are operating within normal parameters.

Recent incident

—

Loading…

—

Core systems

SKIA Frontend —

SKIA Backend / OS —

Database —

Skia-Serve (LLM) —

Image & Video Engines —

Search & Intelligence —

EPAAS Security —

Model & routing

Sovereign models on Skia-Serve (:11500) use a 2M input window. When a task exceeds that band, routing targets skia:llama4-latest with a 10M manifest window. Google Gemini remains continuity fallback only.

Benchmark scores

Benchmark results will appear here after the next intelligence evaluation.

Operational metrics (7d)

Loading aggregated rates…

Capability maturity

Capability updates will appear here as SKIA evolves.

Reasoning engine

Reasoning insights will appear when new analysis is published.

Memory health

Memory health summaries will appear when available.

Intelligence diagnostics

No weakness analysis yet.

⬡ SKIA Intelligence Report

Preparing your intelligence report…