Hermes Agent fails to start after installation is completed, usually not because the installation failed, but because the shell has not yet loaded new commands. The official installation documentation recommends reloading the shell after installation before running setup and main commands.
Three-step inspection
- Execute
source ~/.bashrcfirst, and if you use zsh, executesource ~/.zshrc. - Run the
hermes setupand match the model provider and API key. - Run
hermesto enter the terminal chat interface.
If prompted command not found
Turn off the current terminal and turn it back on, and then try which hermes. If you still can't find it, it means that the step of writing the PATH to the installation script has not taken effect, go back to the installation log to see if there are any permission, network, or dependency errors.
If it can start, but the model will give an error
That's no longer a startup issue, it's a provider/API key/model name issue. Re-run the hermes setup and test it with the most stable remote model.
In a word: the startup order is source shell, hermes setup, and hermes.
Official open source address: https://github.com/NousResearch/hermes-agent; Official document entry: https://hermes-agent.nousresearch.com/.