MyHermes Engineering Blog

Building the next generation of AI agents.

Deep dives into the hardest infrastructure problems in AI, and how MyHermes solves them for you.

2026-06-18Engineering

Why Hermes Runs as a Persistent Private Runtime

A continuously running Hermes container keeps conversations and scheduled work available without provisioning a new environment for every message.

Read Article
2026-06-15Engineering

Bring Your Own Model Provider to Hermes

Use your own OpenRouter, OpenAI, or Anthropic credentials and keep model usage under your control.

Read Article
2026-06-12Engineering

Keeping Agent Credentials Inside the Right Runtime

Tenant-aware routing and private runtime configuration prevent one account from reaching another user's Hermes instance.

Read Article
2026-06-10Engineering

How Durable Hermes Memory Fits Into a Hosted Agent

Hermes stores durable memory in the user's persistent runtime, where it can be inspected through the dashboard.

Read Article
2026-06-05Engineering

Connecting Hermes to Apps Through Composio

Composio provides managed authorization for the app integrations users explicitly enable.

Read Article
2026-06-01Engineering

Seeing Hermes Tool Activity in Web Chat

The web chat streams Hermes responses and exposes tool calls instead of showing only a final block of text.

Read Article
2026-05-28Engineering

Enforcing Hosted Runtime Plan Limits

Resource, channel, app, Kanban, cron, and storage limits are enforced outside the marketing UI.

Read Article
2026-05-20Engineering

Using Hermes From Telegram and WhatsApp

Connect a native messaging channel and continue working with the same private Hermes runtime.

Read Article
2026-05-15Engineering

Turning Agent Requests Into Kanban Work

Hermes can create and execute structured tasks on its built-in Kanban board.

Read Article
2026-05-10Engineering

Scheduling Recurring Work With Hermes Cron Jobs

Create, inspect, pause, resume, and run recurring Hermes jobs from one dashboard.

Read Article