agents

rafaelcalleja's avatarfrom rafaelcalleja

Dynamic agent composition and management system. USE WHEN user says create custom agents, spin up custom agents, specialized agents, OR asks for agent personalities, available traits, agent voices. Handles custom agent creation, personality assignment, voice mapping, and parallel agent orchestration.

⭐0stars🔀0forks📁View on GitHub🕐Updated Jan 2, 2026

When & Why to Use This Skill

The Agents skill is a comprehensive system for dynamic agent composition and management. It enables the creation of specialized AI agents with unique personalities, traits, and voices, facilitating complex multi-agent orchestration and parallel task execution to solve sophisticated problems efficiently.

Use Cases

  • Creating a custom team of specialized agents (e.g., legal experts, security analysts) to perform multi-faceted reviews of documents or projects.
  • Orchestrating parallel agents to handle large-scale data processing or repetitive tasks using different model strengths like Haiku for speed or Opus for reasoning.
  • Developing persistent AI personas with specific backstories and voice mappings for consistent brand communication or specialized domain consulting.
  • Dynamic routing of user requests to the most appropriate agent type based on required expertise, personality traits, or task complexity.
nameAgents
descriptionDynamic agent composition and management system. USE WHEN user says create custom agents, spin up custom agents, specialized agents, OR asks for agent personalities, available traits, agent voices. Handles custom agent creation, personality assignment, voice mapping, and parallel agent orchestration.

Agents - Custom Agent Composition System

Auto-routes when user mentions custom agents, agent creation, or specialized personalities.

Overview

The Agents skill provides complete agent composition and management:

  • Dynamic agent composition from traits (expertise + personality + approach)
  • Personality definitions and voice mappings
  • Custom agent creation with unique voices
  • Parallel agent orchestration patterns

Workflow Routing

Available Workflows:

  • CREATECUSTOMAGENT - Create specialized custom agents -> Workflows/CreateCustomAgent.md
  • LISTTRAITS - Show available agent traits -> Workflows/ListTraits.md
  • SPAWNPARALLEL - Launch parallel agents -> Workflows/SpawnParallelAgents.md

Route Triggers

CRITICAL: The word "custom" is the ABSOLUTE trigger - NO EXCEPTIONS:

User Says What to Use subagent_type Why
"custom agents", "create custom agents", "spin up custom" AgentFactory general-purpose Unique prompts + unique voices
"agents", "launch agents", "bunch of agents" Generic prompt Intern Same voice, parallel grunt work
"use [agent name]", "get [agent name] to" Direct call Named agent type Pre-defined personality

CONSTITUTIONAL RULE FOR CUSTOM AGENTS: When user says "custom agents", you MUST:

  1. Use AgentFactory to compose EACH agent with DIFFERENT traits
  2. Use subagent_type: "general-purpose" - NEVER "Intern", "Designer", "Architect", etc.
  3. Each agent gets their own voice from the trait-to-voice mapping

Architecture

Hybrid Agent Model

Two types of agents:

Type Definition Best For
Named Agents Persistent identities with backstories Recurring work, voice output
Dynamic Agents Task-specific specialists from traits One-off tasks, parallel work

Components

Data/Traits.yaml - 28 composable traits with voice mappings Templates/DynamicAgent.hbs - Agent prompt template Tools/AgentFactory.ts - Dynamic composition engine AgentPersonalities.md - Named agent definitions

Usage

Just ask naturally:

  • "I need a legal expert to review this contract"
  • "Spin up 5 custom science agents"
  • "Get me someone skeptical about security"

The skill routes to appropriate workflow automatically.

Model Selection

Task Type Model Why
Grunt work haiku 10-20x faster
Standard analysis sonnet Balanced
Deep reasoning opus Maximum intelligence