Does it support solo Claude Code sessions?

No, this skill is specifically designed to analyze the interactions and orchestration overhead of multi-agent teams.

How does the skill evaluate my agent team performance?

It fetches the latest official documentation from code.claude.com and evaluates your session against 10 specific categories, including Lead Orchestration, Task Sizing, and Context Sharing.

Can this skill help me save money on API costs?

Yes. One of the primary rubric categories is Model Selection and Cost Efficiency, which identifies if you are over-using high-cost models for simple tasks.

What is required to run a session analysis?

You need a markdown export of an agent team session. You can generate this by using the 'agent-teams:view-team-session' skill with a specific session ID.

Where can I find the generated analysis report?

The analysis is saved as a markdown file in the '.claude/output/' directory of your current project, named after your agent team.

Agent Team Session Analyst

Name: Agent Team Session Analyst
Author: kenneth-liao

bykenneth-liao

•

123

•

협업 도구

Evaluates Claude Code agent team sessions to provide performance scores, actionable optimization advice, and improved prompts based on official best practices.

The Agent Team Session Analyst provides a comprehensive audit of multi-agent workflows within Claude Code. By ingestng session export files, it evaluates team structures, task delegation, and communication patterns against the latest official documentation. Users receive a structured markdown report featuring a suitability verdict, a 10-point scorecard, and evidence-backed recommendations. This skill is essential for developers looking to optimize their agent team usage, reduce token costs, and refine their orchestration prompts for complex engineering tasks.

주요 기능

01Detailed suitability verdict to determine if tasks warrant a multi-agent approach

02123 GitHub stars

03Full prompt refactoring to align with agent team best practices

04Structured markdown report generation saved directly to the workspace output directory

05Automated 10-category rubric evaluation based on official Claude Code documentation

06Evidence-based recommendations citing specific passages from session logs

사용 사례

01Learning best practices for agent orchestration by analyzing real-world session evidence

02Reviewing inefficient agent team runs to identify communication bottlenecks or task-sizing issues

03Optimizing token usage and costs for complex multi-agent engineering workflows

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add kenneth-liao/ai-launchpad-marketplace analyze-team-session

For use in Claude.ai and ChatGPT

Download Skill