@difyz/n8n-nodes-hermes-agent
v0.1.2
Published
n8n community node for connecting to a self-hosted Hermes Agent API server
Maintainers
Readme
n8n-nodes-hermes-agent
An n8n community node for connecting workflows to a self-hosted Hermes Agent API server.
Features
- OpenAI-compatible chat completions and Responses API
- Stateful conversations using
previous_response_idor a named conversation - Long-running Hermes runs, polling, stopping, and approval decisions
- Jobs and sessions management
- Models, capabilities, skills, toolsets, and health discovery
- Custom API requests for endpoints added by future Hermes releases
- Can be used as an n8n AI tool
Streaming endpoints are intentionally invoked in non-streaming/polling form because normal n8n node executions produce completed items rather than an open SSE stream.
Install
Install n8n-nodes-hermes-agent from Settings → Community Nodes, or install it into a self-hosted n8n installation with npm.
For local development:
npm install
npm run devConfigure Hermes
Add the following to the Hermes profile's .env, then start the gateway:
API_SERVER_ENABLED=true
API_SERVER_KEY=replace-with-a-strong-secrethermes gatewayCreate an Hermes Agent API credential in n8n with:
- Base URL: for example
http://host.docker.internal:8642when n8n runs in Docker and Hermes runs on the host - API Key: the value of
API_SERVER_KEY
Do not use localhost from an n8n container unless Hermes runs in the same container. The API grants access to Hermes' full toolset, so expose it only on trusted networks and protect the key.
Development
npm run lint
npm run buildLicense
MIT
