Hermes Agent Interactive Configuration Guide

Use the built-in hermes model wizard to connect to Mai. The wizard is responsible for writing to ~/.hermes/config.yaml and ~/.hermes/.env, suitable for users who do not want to manually edit YAML.

Configuration Steps

  1. Create a Coding-specific API Token in the Mai console.
  2. Run hermes model, and choose to add a custom OpenAI-compatible Provider.
  3. Fill in the following:
Field Value
Provider name acedatacloud
Base URL https://api.ai.moses.day/v1
API key environment variable ACEDATACLOUD_API_KEY
Model The model ID selected in the Coding configuration wizard
API format Chat Completions
  1. After saving, completely exit and restart Hermes.

The token is only written to the local credentials file; do not submit it to repositories, issues, logs, or screenshots.

Verification

hermes --version
hermes chat --model MODEL_ID "Reply only OK"

If it returns 401, check the Token in ~/.hermes/.env; if changes do not take effect, restart Hermes. Some services that only provide streaming responses also need to be registered in Hermes's auxiliary.stream_only_base_urls, but this configuration for regular chat does not require additional settings.

Official References

Return to Coding Plan Configuration Center.