team-builder

Discovers agent personas and dispatches selected agents in parallel for task execution.

Updated Mar 18, 2026
One-click install
npx skills add https://github.com/freedom909/real-estate-saas --skill team-builder-freedom909
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: team-builder
Source: https://github.com/freedom909/real-estate-saas/tree/main/.trae/skills/team-builder
Command: npx skills add https://github.com/freedom909/real-estate-saas --skill team-builder-freedom909

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Choosing and coordinating multiple AI agent personas for a task is manual and error-prone. This Skill automates agent discovery, selection, and parallel dispatch so you can assemble ad-hoc expert teams on demand. ## Core Features & Use Cases - Dynamic Agent Discovery: Scans claude agents output and markdown files in project or global agent directories, supporting both flat and domain-subdirectory layouts. - Interactive Team Composition: Presents a domain-grouped menu and accepts flexible selection by numbers, names, or whole domains, capped at 5 agents. - Parallel Dispatch & Synthesis: Spawns all selected agents simultaneously via the Agent tool and merges their outputs into a unified report highlighting agreements, conflicts, and next steps. - Use Case: Before launching a Next.js e-commerce site, select a Security Engineer and SEO Specialist agent, have both review the codebase in parallel, and receive a synthesized report reconciling their recommendations. ## Quick Start Ask the assistant to run team builder and pick agents to review your project.

Frequently Asked Questions about team-builder

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run multiple AI agents in parallel on one task?▼

Select up to five agents from the discovered menu, provide a task description, and the Skill spawns all agents simultaneously using parallel Agent tool calls. Each agent runs independently and results are merged into one report.

How are agent personas discovered from markdown files?▼

Discovery merges the claude agents command output with markdown files found in ./agents and ~/.claude/agents directories. The first heading becomes the agent name and the first paragraph becomes its description, with duplicates resolved by source precedence.

Does team-builder support flat and subdirectory agent layouts?▼

Yes, both layouts are supported. Subdirectory layouts infer the domain from the folder name, while flat layouts infer domains from filename prefixes shared by two or more files, grouping unique prefixes under General.

What is the maximum number of agents per team?▼

The limit is five agents per team, enforced at selection time. If you select more, the Skill lists them alphabetically and asks you to narrow down or keep the first five.

What happens if one of the parallel agents fails?▼

The failure is noted inline with the reason, such as an error, timeout, or empty output, and the Skill continues synthesizing results from the agents that succeeded.