agent-expenditure

This Skill addresses the challenge of tracking token consumption and mitigating cost waste from excessive coordination overhead in multi-agent parallel scheduling scenarios. In practice, developers often need to deploy multiple AI agents concurrently to handle complex tasks, but they lack effective tools to quantify each agent's contribution and consumption. This can lead to inefficient 'brute-force' approaches and unnecessary resource waste.

Usage is straightforward: after completing a multi-agent parallel scheduling task, simply trigger this Skill for analysis. You can then use natural language instructions to obtain a detailed token usage report, including consumption per agent, the proportion of coordination overhead (e.g., file read/write conflicts), and signals of waste such as duplicated work.

It is particularly suitable for teams or individuals coordinating multiple agents for complex tasks (e.g., large-scale code reviews, parallel data processing, multi-perspective competitive analysis). This is especially valuable for those already using parallel scheduling strategies but concerned about cost overruns, or for developers wanting to evaluate whether adding more agents genuinely improves efficiency.

We recommend using this tool for post-task review after each parallel operation involving four or more agents. Note that it is primarily a post-analysis tool and is not designed for single-agent workflows or real-time token budget control, which should be handled by dedicated token budget management tools.