MC-1 Core

Overview

MC-1 separates an application's intelligence intent from the infrastructure that executes it. It filters ineligible paths first, then optimizes among eligible paths and records evidence.

• One logical model: colomboai/mc-1.
• Policy-aware eligibility before optimization.
• Objectives for quality, cost, latency, privacy, sovereignty, safety, and reliability.
• Evaluation and bounded escalation.
• Provider and model attribution remains visible.

Production checklist

• Keep credentials in a server-side secret store.
• Set explicit cost, latency, privacy, provider, and regional constraints.
• Capture route and evaluation evidence for incident review.
• Test timeouts, cancellation, fallback, and denied-policy paths before rollout.