agent-platform

Build an LLM multi-agent platform that delegates tasks and routes SSE responses.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/escotilha/claude-public --skill agent-platform
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: agent-platform
Source: https://github.com/escotilha/claude-public/tree/main/skills/_archive/agent-platform
Command: npx skills add https://github.com/escotilha/claude-public --skill agent-platform

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you build an LLM multi-agent platform that reliably delegates work to specialist agents while routing their responses back to the right chat channels.

Core Features & Use Cases

  • COO orchestration pattern: A primary orchestrator agent delegates to specialist agents (researcher, writer, ops, or dynamically created agents) and returns a short summary to the COO channel.
  • SSE-based routing: Streams “delegation:task”, “delegation:response”, and “agent:created” events so each specialist’s full response appears in their own channel.
  • Dynamic agent creation: Creates new agents with persisted personas, seeds default capabilities, and updates the UI sidebar without reload.
  • Practical production safeguards: Includes compatibility fixes (e.g., Safari cache-control for API/auth routes), tooling-call behavior guidance, and pitfalls for DB/migration mismatches.

Quick Start

Ask your orchestrator to create a new specialist agent and delegate the user’s request so the specialist’s streamed response routes back to the specialist’s channel.

Frequently Asked Questions about agent-platform

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

FAQPage Schema
How do I build an LLM multi-agent orchestration platform for task delegation?▼

To build multi-agent orchestration, you need a primary orchestrator agent that delegates user requests to specialist agents and routes their streamed responses back to the correct chat channels using Server-Sent Events (SSE).

How does SSE routing work for multi-agent chat experiences?▼

SSE routing works by broadcasting specific event types like delegation:task, delegation:response, and agent:created, ensuring each specialist agent's full streamed response appears in its own designated UI channel without requiring a page reload.

What is the COO orchestration pattern for LLM workflows?▼

The COO orchestration pattern is a delegation workflow where a primary orchestrator agent assigns tasks to specialist agents like a researcher or writer, then returns a concise summary of the results back to the main COO channel.

How do I dynamically create specialist agents with persisted personas?▼

You can dynamically create specialist agents by dispatching a delegate call that generates the new agent, assigns a persisted persona, seeds default capabilities, and immediately updates the UI sidebar to reflect the new specialist.

What production safeguards are needed for multi-agent SSE routing?▼

Production safeguards for SSE routing include applying Safari cache-control fixes for API and auth routes, validating tooling-call behaviors, and implementing guardrails to prevent database and migration mismatches.

Can I use dynamic agent onboarding without reloading the chat dashboard?▼

Yes, dynamic agent onboarding updates the UI sidebar automatically without a reload by broadcasting an agent:created SSE event that immediately integrates the new specialist into the existing chat dashboard.