Skill UI

Cavecrew provides a set of specialized subagents (investigator, builder, reviewer) designed for efficient code development delegation. Unlike vanilla agents, Cavecrew compresses output results, drastically reducing the token usage injected back into the main context. This allows users to maintain context over much longer, complex coding sessions, effectively preventing context exhaustion during large-scale refactoring, debugging, or multi-step feature development.

This advanced skill manages complex, large-scale development tasks by coordinating multiple, specialized subagents in parallel. It systematically decomposes the goal into isolated missions, manages resource quotas (token usage), and handles error recovery across multiple files and components. Ideal for large feature builds, system refactoring, or complex applications requiring multiple, concurrent roles (Planner, Builder, Tester).

A powerful tool for designing and writing deterministic, multi-agent workflow scripts in JavaScript. It allows users to orchestrate complex, repeatable, multi-step tasks using advanced patterns like fan-out, pipelines, loops, and judge-panels. Ideal for automating sophisticated business processes that require multiple specialized sub-agents and robust context management.

Mason is the dedicated builder agent responsible for writing clean, functional, and production-ready code. He strictly adheres to architectural blueprints and defined checklists, ensuring every feature is implemented with high code quality standards, including comprehensive error handling, security best practices, and testability. Ideal for the core development phase of complex projects.

This skill runs an automated evolution process designed to systematically optimize the seven core policy surfaces of an AI agent harness (e.g., planner, contextBuilder, reviewer). By mutating these policies and scoring the variants in a sandbox, it identifies crucial performance improvements without requiring expensive foundation model retraining. It is ideal for empirical configuration tuning and establishing optimized baselines.

An interactive tool designed to discover, select, and assemble specialized teams of AI agents. It supports structured agent collections across different domains (e.g., Engineering, Marketing). Users select multiple roles for a specific task, and the skill launches them in parallel, synthesizing their independent findings into a unified report, including consensus and conflict analysis.

This skill summarizes user-provided input into one concise paragraph. It is specifically designed to enhance E2E (End-to-End) coverage testing within development environments, such as Agent Builders. The generated summary ensures comprehensive content representation and must include the E2E_SUMMARY_SKILL_PASS marker for validation.