runaway-guard

What problem does it solve?

This Skill prevents unexpected AI inference bills by forcing developers to define cost limits, retry bounds, and provider safeguards before writing paid API integrations.

Core Features & Use Cases

• Cost Safety Contracts: Establishes explicit per-run costs, daily spending caps, maximum calls, and concurrency limits for paid AI API workflows.
• Runaway Prevention Controls: Audits loops, retries, fan-out pipelines, agent flows, polling jobs, and webhook paths for unbounded spending risks.
• Use Case: Apply this Skill when building an AI image generation pipeline or agent workflow that calls paid models and needs protection against accidental retry storms or excessive API usage.

Quick Start

Use the runaway-guard skill to review my paid AI API integration and define the required cost contract, limits, and safeguards before implementation.