token-cap

Enforce token usage ceilings per route in AI dispatch harnesses.

Updated May 11, 2026
One-click install
npx skills add https://github.com/AesopScott/mojo --skill token-cap
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: token-cap
Source: https://github.com/AesopScott/mojo/tree/main/harnesses/skills/token-cap
Command: npx skills add https://github.com/AesopScott/mojo --skill token-cap

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill prevents runaway costs and resource exhaustion by enforcing strict token ceilings on specific API routes, ensuring that individual dispatches cannot exceed defined usage boundaries.

Core Features & Use Cases

  • Cost Efficiency Control: Sets hard limits on token consumption to prevent overspending during model inference.
  • Harness Boundary Management: Provides a structured runbook for reviewing and adjusting route configurations, policies, and dispatch logic.
  • Use Case: If a specific route is consistently exceeding its budget, use this Skill to define a new token cap, implement the change, and verify that the system correctly blocks or throttles requests that exceed the limit.

Quick Start

Use the token-cap skill to inspect the current route configuration and propose a new token ceiling that prevents overspending on the target endpoint.

Frequently Asked Questions about token-cap

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

FAQPage Schema
How do I enforce token limits per route to prevent API cost overruns?

To enforce token limits per route, you apply strict token ceilings to specific API routes using configuration files and policy definitions. This blocks or throttles dispatches that exceed the defined token budget, maintaining cost efficiency and preventing resource exhaustion.

What is token cap policy in AI dispatch harnesses?

A token cap policy in AI dispatch harnesses is a strict boundary on token expenditure for specific routes. It ensures individual model inference dispatches cannot exceed defined usage limits, preventing runaway costs and resource overconsumption.

How do I set up route management to throttle requests exceeding a token budget?

Set up route management by inspecting current route configurations and proposing a new token ceiling. The system then validates boundary enforcement against observability logs, blocking or throttling requests that exceed the defined limit.

Do I need observability logs to validate token boundary enforcement?

Yes, you need observability logs to validate token boundary enforcement. The system requires access to route configuration files, policy definitions, and observability logs to verify that token ceilings are correctly applied and maintained.

What is the best way to control model inference costs on specific API routes?

The best way to control model inference costs is to set hard limits on token consumption per route. This provides structured harness boundary management, ensuring individual dispatches cannot exceed their defined usage boundaries.

Why does a specific route consistently exceed its token budget?

A route exceeds its token budget when it lacks a strict token ceiling. Implement a new token cap in the route configuration and verify the system correctly blocks or throttles requests to prevent overspending on the target endpoint.