llm-lab: deployment tool for Hermes Agent

llm-lab

tecno-consultores/llm-lab

Docker Compose stack with profiles for n8n, Open WebUI, Ollama, Qdrant, OpenCode and Hermes Agent

In short

llm-lab is a Docker Compose project from Tecno Consultores in which n8n, Open WebUI, Ollama, Qdrant, OpenCode, Hermes Agent and other self-hosted AI services are individual Compose profiles you switch on as needed.

What llm-lab does

Each service in llm-lab is a Compose profile. The main table lists n8n with Postgres and KeyDB, n8n worker, runner and AI-sandbox variants, Open WebUI, Ollama for Nvidia, AMD or CPU, EvolutionAPI, OpenCode and Hermes-agent. Hermes runs on both AMD64 and ARM64 and starts with --profile hermes. It uses ports 8642 and 9119, the dashboard is at http://127.0.0.1:9119, and inside Docker the OpenAI-compatible API is at http://hermes:8642/v1.

Complementary profiles add Searxng, Browserless, NGINX Proxy Manager, Postgres, KeyDB, OmniRoute, Qdrant and a Hermes-opencode-bridge service that depends on the hermes and opencode profiles and listens on port 8000. The services have healthchecks, and the README gives one command to stop and remove all containers. It also warns that PostgreSQL 18 changes the database directory structure, which matters when upgrading from earlier versions.

Key features

One Compose file with a profile per service
Hermes-agent profile for AMD64 and ARM64 with dashboard on port 9119 and API on port 8642
Ollama profiles for Nvidia GPU, AMD GPU or CPU
Companion services such as Qdrant, Searxng, Browserless and OmniRoute
Healthchecks and a table of ports and URLs for every service

When to use it

Running Hermes Agent next to n8n and Open WebUI on one Docker host
Giving n8n workflows an OpenAI-compatible Hermes endpoint inside Docker
Pairing Hermes with OpenCode through the bridge profile

Who it is for: Self-hosters who want a Docker Compose lab that includes Hermes Agent among other AI and automation services.

How it fits with Hermes Agent

Supports Hermes Agent among many services, as the hermes profile plus an optional Hermes-opencode-bridge.

Requirements: Docker with Compose; an Nvidia or AMD GPU only for the GPU Ollama profiles

Note: The sample command in the README starts n8n and Open WebUI, so Hermes needs --profile hermes added.

FAQ

What is llm-lab?

llm-lab is a Docker Compose project that bundles self-hosted AI and automation services as profiles. It includes n8n, Open WebUI, Ollama, Qdrant, EvolutionAPI, OpenCode and Hermes Agent.

How do I start Hermes Agent with llm-lab?

Add --profile hermes to the docker compose command that uses docker-compose.yml and env.example. The dashboard is then at http://127.0.0.1:9119, and the API is at http://hermes:8642/v1 inside Docker.

Is llm-lab free and open source?

Yes, the repository is released under the GPL-2.0 license.