Hermes Agent / Agent Plan reference

Hermes Agent integration reference

Configure Hermes with a separate YAML route and environment secret while keeping the model selection easy to audit.

1 · Quick start

Connect Hermes Agent in three moves

Start with the matching configuration file, keep the key in the named secret location, and use one exact Model ID throughout the setup.

01

Create the Hermes files

Keep the route in ~/.hermes/config.yaml and the key in ~/.hermes/.env. Do not paste the raw key into YAML or commit either file.

02

Use the custom endpoint

Set the provider base URL to the /v3 OpenAI-compatible route and tell Hermes to read the key from ARK_API_KEY.

03

Choose the default model

Set model.default to one exact Model ID from the Agent Plan page. Keep the value unchanged while diagnosing connectivity.

The key is delivered to your Dashboard after payment. Replace the placeholder in the example, then keep the real value in your environment or local secret file.

2 · Configuration

Copy the provider blueprint

This is a safe starting point for the current Agent Plan route. Keep existing tool settings that are unrelated to the provider and replace only the placeholder key and Model ID values.

# ~/.hermes/.env
ARK_API_KEY=YOUR_AGENT_PLAN_KEY

# ~/.hermes/config.yaml
model:
  default: glm-5.2
provider:
  base_url: https://ark.cn-beijing.volces.com/api/plan/v3
  api_key_env: ARK_API_KEY

The default Model ID shown here is the stocked plan's current default. If you choose another supported model, replace every model field required by Hermes Agent, not just the visible label.

Install / launch

Install Hermes Agent using the release instructions for your platform.

3 · Verify the route

Confirm the connection before a full project

Restart Hermes after changing either the .env file or config.yaml.

Run a short prompt and confirm the response before connecting tools or long-running tasks.

If your installed Hermes version uses different YAML field names, keep the same three values: key environment name, /v3 base URL, and Model ID.

Troubleshooting

Common configuration pitfalls

Do not place the key directly in a public config file or shell history.

The custom endpoint is OpenAI-compatible; the Claude Code base URL is not interchangeable with it.