Discord Setup Issues

[pine mulch]

I am using Hermes on discord, there's a server with only the two of us . I have created multiple channels for specific areas and niches.. can anyone tell me how I can set different models to work on that particular channel? It was very easy on openclaw i just had to mention it in discord chat but it's not the same case here. Also I have turned reply_to_message: false , but it still replies to message (like you now hold and click reply instead of just typing normally) i have asked it multiple times can't seem to get this sorted ..

Also steps to get the web UI working for Hermes please

Should I have a multi agent setup for each channel/category (does that allow choosing different models for specific areas?)

Neither me nor Hermes itself can fix the reply problem

[kindred timber]

Yes, for different models per Discord area, use a multi-profile / multi-gateway setup.

Hermes profiles are isolated. Each profile has its own config, provider/model settings, memory, sessions, skills, and gateway state. So the clean setup is:

research profile → research Discord bot token → research channels → research model
coding profile → coding Discord bot token → coding channels → coding model
writing profile → writing Discord bot token → writing channels → writing model

You will need separate Discord bot tokens/keys for the separate active gateways. Do not run multiple active gateways against the same Discord bot token.

Then restrict each bot/profile to the channels it should answer in with discord.allowed_channels in that profile’s config.yaml, or DISCORD_ALLOWED_CHANNELS in that profile’s .env.

Example:

discord:
  allowed_channels:
    - 123456789012345678
    - 987654321098765432

Hermes also supports discord.channel_prompts, but that only changes instructions/personality per channel. It does not switch the underlying model. For actual different models/providers per area, profiles + separate gateways is the right path.

For the reply-reference issue, reply_to_message: false is not the Hermes Discord setting. Use this instead:

discord:
  reply_to_mode: "off"

Or in .env(legacy setting):

DISCORD_REPLY_TO_MODE=off

Then restart the gateway. If DISCORD_REPLY_TO_MODE is set in .env, it wins over config.yaml.

Be warned. Disabling this for multiple bots in one channel will cause infinite loops.

You must restart the gateway for these changes to take effect.

For the web UI, run:

hermes dashboard

Then open:

http://127.0.0.1:9119

Docs:

Discord:
https://hermes-agent.nousresearch.com/docs/user-guide/messaging/discord

Profiles:
https://hermes-agent.nousresearch.com/docs/user-guide/profiles

Web dashboard:
https://hermes-agent.nousresearch.com/docs/user-guide/features/web-dashboard

Docker dashboard:
https://hermes-agent.nousresearch.com/docs/user-guide/docker#running-the-dashboard