otp-agent-orchestration

Name: otp-agent-orchestration
Availability: InStock
Author: phrazzld

Orchestrate concurrent Elixir/OTP AI agent processes with DynamicSupervisor and webhook dispatch.

13|3|Updated Mar 2, 2026

Search Tags:#concurrency #telemetry #webhook #elixir #otp #agent-orchestration #supervision

Ready to run

npx skills add https://github.com/phrazzld/spellbook --skill otp-agent-orchestration

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: otp-agent-orchestration
Source: https://github.com/phrazzld/spellbook/tree/main/skills/otp-agent-orchestration
Command: npx skills add https://github.com/phrazzld/spellbook --skill otp-agent-orchestration

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Elixir/OTP patterns for concurrent AI agent orchestration services. Use when designing systems that manage many concurrent, long-running LLM agent processes with fault tolerance, rate limiting, and webhook-driven dispatch.

Core Features & Use Cases

Dynamic supervision for agent pools and recovery across failures
GenServer-based rate limiting and per-model throttling
Phoenix webhook ingestion and Port-based CLI dispatch for external triggers
ETS-based registries for active runs and cooldown tracking
Telemetry for observability and cost tracking
Use cases include deploying an orchestrator to run dozens to hundreds of agents concurrently

Quick Start

Start the OTP-based orchestrator to supervise and dispatch concurrent AI agents via webhook-driven events.

Frequently Asked Questions about otp-agent-orchestration

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

FAQPage Schema

How do I coordinate concurrent AI agents using Elixir OTP?▼

This orchestrator coordinates concurrent AI agents using Elixir OTP by leveraging DynamicSupervisor for agent pools, GenServer for rate limiting, and Phoenix webhooks for dispatch to manage tens to hundreds of fault-tolerant processes.

How do I run long-running LLM workflows without losing data during failures?▼

To run long-running LLM workflows without losing data during failures, you apply OTP supervision trees with DynamicSupervisor and ETS registries, enabling automatic recovery and tracking for active agent runs across crash events.

What is the best way to rate limit LLM API requests per model in Elixir?▼

The best way to rate limit LLM API requests per model in Elixir is using GenServer-based throttling mechanisms, which provide per-model rate limiting alongside ETS registries for cooldown tracking to prevent exceeding API limits.

Does this Elixir OTP agent orchestration approach scale to hundreds of concurrent agents?▼

Yes, Elixir OTP agent orchestration scales to hundreds of concurrent agents by utilizing DynamicSupervisor for scalable pool management and Port-based CLI dispatch to handle high-volume webhook-driven workloads efficiently.

How do I trigger AI agent processes via webhooks in an Elixir application?▼

You trigger AI agent processes via webhooks in an Elixir application by exposing a Phoenix Endpoint for ingestion, which dispatches events to DynamicSupervisor pools using Port-based CLI execution for external triggers.

How do I track costs and observability for concurrent LLM agent runs?▼

You track costs and observability for concurrent LLM agent runs by implementing Elixir telemetry alongside ETS-based registries, enabling detailed monitoring of active runs and resource usage across the supervised agent pool.