rate-limit-management

Enforce three-tier wind-down protocols for Claude Max usage thresholds.

88|108|Updated Apr 7, 2026
One-click install
npx skills add https://github.com/grandamenium/cortextos --skill rate-limit-management-grandamenium
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: rate-limit-management
Source: https://github.com/grandamenium/cortextos/tree/main/community/skills/rate-limit-management
Command: npx skills add https://github.com/grandamenium/cortextos --skill rate-limit-management-grandamenium

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Prevents Claude Code agents from crashing, corrupting in-progress work, or triggering hard API failures when Claude Max usage thresholds are reached.

Core Features & Use Cases

  • Three-tier wind-down protocol: Defines Wind-Down (>= 85%), Minimal (85-95%), and Dark (>= 95%) behaviors to reduce load safely while preserving user-facing responsiveness.
  • Deterministic usage check command: Provides a single command to query Anthropic OAuth usage via cortextos bus check-usage-api, including early-warning thresholds for proactive Telegram alerts.
  • Automatic recovery rules: Resumes normal operations only after utilization drops below 20%, restoring suspended cron work and notifying the user.

Quick Start

Run the usage check with early warnings by executing: cortextos bus check-usage-api --warn-7day 75 --warn-5h 85 --chat-id $CTX_TELEGRAM_CHAT_ID.

Frequently Asked Questions about rate-limit-management

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I prevent Claude API failures when agents hit Max usage limits?

Prevent Claude API failures by enforcing a three-tier wind-down protocol that suspends or throttles agent behavior when Claude Max usage reaches 85% and 95% thresholds, ensuring long-running operations pause safely without crashing.

How does the three-tier wind-down protocol manage agent throttling?

The three-tier wind-down protocol manages agent throttling by applying Wind-Down behavior at 85% utilization, Minimal behavior between 85-95%, and Dark behavior above 95%, progressively reducing load while preserving responsiveness and automatically resuming below 20%.

Can I monitor Claude API utilization and send proactive Telegram alerts?

Monitor Claude API utilization and send proactive Telegram alerts by executing the deterministic check-usage-api command with custom warning thresholds for 7-day and 5-hour rolling windows, configured directly to your Telegram chat ID.

What's the best way to handle cron job suspension during high API utilization?

Handle cron job suspension during high API utilization by applying automatic response and cron suspension rules when rolling 5-hour and 7-day usage thresholds are exceeded, restoring suspended work only after utilization drops below 20%.

Does this rate limiting approach work for scheduled multi-agent operations?

This rate limiting approach works specifically for scheduled, proactive, and long-running multi-agent operations, applying cached usage retrieval and throttling rules to safely pause work based on rolling 5-hour and 7-day Claude Max usage metrics.

Why do long-running proactive agents crash and how do I recover them?

Long-running proactive agents crash when Claude Max usage thresholds are breached, triggering hard API failures; recovery occurs automatically when utilization drops below 20%, restoring suspended cron work and notifying the user.