resource-exhaustion_skill

Overview

This skill teaches how to find, understand, and mitigate resource exhaustion and denial-of-service (DoS) vulnerabilities in AI-generated JavaScript code. It focuses on common failure modes like unbounded loops, missing rate limits, unchecked uploads, and expensive API usage. You’ll get practical controls, patterns, and code-level defenses to prevent crashes and runaway costs.

How this skill works

The skill inspects typical AI-generated patterns and highlights where functionality-first code can exhaust CPU, memory, network, or API budget. It explains concrete mitigations: input validation, size/time limits, rate limiting, job queues, whitelists, and fail-safe timeouts. Code snippets and configuration examples show how to apply limits and replace dangerous patterns with controlled implementations.

When to use it

• When reviewing AI-generated endpoints that process files, images, or large payloads
• When designing features that call costly third-party AI APIs
• When hardening upload, batch-processing, or recursive code paths against abuse
• When implementing rate limiting, queuing, or concurrency controls
• When assessing production cost risk from unbounded user input

Best practices

• Validate and whitelist inputs: URLs, file types, operation names, and sizes before processing
• Set strict resource limits: max file size, max dimensions, max operations, timeouts, and concurrency caps
• Use rate limiting and per-user quotas to prevent mass abuse and API cost spikes
• Queue expensive jobs and limit concurrent workers; return job IDs instead of processing synchronously
• Apply defensive timeouts and abort controllers for network calls and long-running operations
• Monitor usage and set automatic spend alerts and API usage caps for third-party services

Example use cases

• Secure an image-processing endpoint: reject oversized images, restrict operations, and queue work
• Protect an AI summarization API: enforce text length, per-user daily limits, and shorter model/timeouts
• Harden file upload routes: validate MIME types and reject files over configured size limits
• Prevent DB overload: add pagination, max limits, and indexed queries before returning results
• Mitigate recursive rendering: limit recursion depth or convert to iterative traversal before rendering