How does this skill help reduce AI development costs?

It identifies 'waste signals' such as redundant work or high coordination overhead, allowing you to optimize agent counts and avoid unnecessary token consumption in future runs.

When is the best time to use the Agent Expenditure Tracker?

It is most effective after parallel agent runs complete or during retrospectives to evaluate whether the expenditure was proportional to the value delivered.

Can this skill prevent token waste in real-time?

While it focuses on post-hoc analysis, it integrates with planning rules that trigger before large dispatches to prevent wasteful execution based on historical insights.

What is Brooks's Law for AI agents?

It is the principle that adding more agents to a task increases coordination overhead (like file conflicts), which can become counterproductive. This skill helps you find the 'sweet spot' for agent count.

Agent Expenditure Tracker

Name: Agent Expenditure Tracker
Author: athola

byathola

•

262

•

分析与监控

Monitors token usage and coordination overhead in multi-agent workflows to optimize resource efficiency.

The Agent Expenditure Tracker provides specialized tools for monitoring and evaluating token consumption during parallel agent dispatches. By applying Brooks's Law to AI workflows, it helps developers identify coordination overhead, detect waste signals like redundant work or shared-file conflicts, and ensure that increasing agent counts actually translates to proportional value. This skill is essential for teams using Claude Code in complex, multi-agent environments where cost-efficiency and output quality must be balanced through post-hoc analysis and architectural retrospectives.

主要功能

01Post-dispatch review checklist to identify redundant work patterns

02Coordination overhead analysis based on Brooks's Law principles

03Detection of waste signals and concurrent file access conflicts

04Per-agent token usage tracking for parallel dispatch workflows

05262 GitHub stars

06Integration with planning rules for high-agent-count tasks

使用场景

01Evaluating the cost-effectiveness of large-scale parallel code generation

02Conducting retrospectives on resource-heavy automated workflows

03Determining the optimal number of agents for specific engineering tasks

主要功能

01Post-dispatch review checklist to identify redundant work patterns

02Coordination overhead analysis based on Brooks's Law principles

03Detection of waste signals and concurrent file access conflicts

04Per-agent token usage tracking for parallel dispatch workflows

05262 GitHub stars

06Integration with planning rules for high-agent-count tasks

使用场景

01Evaluating the cost-effectiveness of large-scale parallel code generation

02Conducting retrospectives on resource-heavy automated workflows

03Determining the optimal number of agents for specific engineering tasks