deployment-helper_skill

Overview

This skill helps deploy elizaOS agents to production with practical patterns for single-agent, multi-agent, and containerized setups. It bundles guidance for health checks, graceful shutdowns, metrics, and common production checklist items. Use it to get a reliable, observable, and scalable deployment pipeline for TypeScript-based elizaOS agents.

How this skill works

The skill inspects runtime configuration and provides runnable patterns: a minimal single-agent runtime, Docker and docker-compose manifests, and a coordinator for multi-agent deployments with simple load distribution. It also integrates Prometheus-style metrics collection endpoints and shows how to implement health checks and graceful shutdown. Finally, it lists a production checklist covering migrations, monitoring, secrets, backups, and rollback planning.

When to use it

• Deploy a single elizaOS agent to production with a minimal footprint.
• Run multiple agents on one host or coordinate per-request routing and basic load distribution.
• Containerize agents with Docker and orchestrate dependencies like Postgres and Redis.
• Add observability: expose /metrics and health endpoints for monitoring and readiness checks.
• Prepare for production readiness audits and operational handoffs.

Best practices

• Expose a /health endpoint and implement graceful shutdown on SIGTERM to avoid in-flight truncation.
• Use a managed database and run migrations before startup; keep connection strings in environment variables.
• Collect metrics via prom-client and expose /metrics for scraping by Prometheus.
• Run agents in Docker with restart policies and separate services for DB and cache; persist volumes for stateful services.
• Implement logging, error tracking (e.g., Sentry), rate limiting, HTTPS, secrets management, and a tested rollback plan.

Example use cases

• Single-agent API for a game asset generation microservice behind a load balancer.
• Docker-compose local staging environment with Postgres and Redis to mirror production.
• Multi-agent coordinator that shards requests across character-specific runtimes for parallelism.
• Add Prometheus metrics to track messages processed and response latency for SLOs.
• Operational checklist automation to validate environment variables, migrations, and backups before deploy.