compute-adaptation

What problem does it solve?

Graceful degradation is needed when compute resources are constrained; this Skill provides structured strategies to keep agent workflows operating instead of failing when limits are encountered.

Core Features & Use Cases

• Four-tier compute model (NORMAL, LOW, CRITICAL, OFFLINE) with defined capabilities.
• Tiered policies and explicit capability flags to govern actions at each tier.
• Recovery and upgrade paths to regain full functionality when resources return.
• Use cases include scaling agent teams under budget limits, handling rate-limiting scenarios, and offline queuing for later execution.

Quick Start

Switch to an appropriate compute tier when limits are reached and automatically apply the degradation strategy.