Is this skill compatible with Cursor and Windsurf?

While optimized for Claude Code, it is part of the Task Orchestrator MCP server which provides context persistence and orchestration capabilities for Claude, Cursor, and Windsurf.

Does this skill block my execution if an error is found?

No, it follows an 'observe and validate' philosophy. It reports issues using severity levels like ALERT and WARN but does not block the workflow, allowing the user to make the final decision.

What is Orchestration QA?

Orchestration QA is a quality assurance skill for Claude that validates if AI agents and sub-skills are following established patterns and workflows accurately during complex tasks.

How does it help with token optimization?

It identifies token waste patterns such as overly verbose output, unnecessary file loading, and redundant tool operations, suggesting more efficient paths for the AI to take.

Can I choose which analyses are active?

Yes, the skill includes an interactive configuration phase where you can enable or disable specific categories like Information Density, Execution Graphs, or Tag Coverage.

Orchestration QA

Name: Orchestration QA
Author: jpicklyk

byjpicklyk

•

128

•

セキュリティとテスト

Validates and improves AI orchestration workflows by monitoring skill adherence, token efficiency, and execution quality.

Orchestration QA is a specialized skill designed for developers using complex AI agent workflows within the Task Orchestrator ecosystem. It ensures that Claude's subagents and skills follow documented patterns, preventing context drift and ensuring high-quality outputs. By providing pre-execution validation and post-execution reviews, it identifies process deviations, optimizes token usage, and suggests continuous improvements for task orchestration, making it indispensable for maintaining reliability and cost-efficiency in automated coding environments.

主な機能

01Pre-execution context capture and checkpoint setting to define expected workflow outcomes

02Interactive configuration to selectively enable specific analysis categories and control token costs

03128 GitHub stars

04Structured deviation reporting using ALERT, WARN, and INFO severity levels for transparent auditing

05Token optimization analysis to identify waste patterns and improve tool selection efficiency

06Post-execution reviews focusing on information density, routing adherence, and output quality

ユースケース

01Ensuring subagents follow strict routing rules and don't bypass critical status progression steps

02Validating complex feature architectures and execution graphs before starting implementation

03Monitoring token efficiency and cost-to-performance ratios across long-running AI sessions

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jpicklyk/task-orchestrator orchestration-qa

For use in Claude.ai and ChatGPT

Download Skill