feat: warn users when loading non-agentic Hermes LLM models (#5378)

Nous Research Hermes 3 & 4 models lack tool-calling capabilities and
are not suitable for agent workflows. Add a warning that fires in two
places:

- /model switch (CLI + gateway) via model_switch.py warning_message
- CLI session startup banner when the configured model contains 'hermes'

Both paths suggest switching to an agentic model (Claude, GPT, Gemini,
DeepSeek, etc.).
This commit is contained in:
Teknium 2026-04-05 18:41:03 -07:00 committed by GitHub
parent e9ddfee4fd
commit fc15f56fc4
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
2 changed files with 44 additions and 1 deletions

16
cli.py
View file

@ -2358,6 +2358,22 @@ class HermesCLI:
"[dim] Fix: Set model.context_length in config.yaml, or increase your server's context setting[/]"
)
# Warn if the configured model is a Nous Hermes LLM (not agentic)
model_name = getattr(self, "model", "") or ""
if "hermes" in model_name.lower():
self.console.print()
self.console.print(
"[bold yellow]⚠ Nous Research Hermes 3 & 4 models are NOT agentic and are not "
"designed for use with Hermes Agent.[/]"
)
self.console.print(
"[dim] They lack tool-calling capabilities required for agent workflows. "
"Consider using an agentic model (Claude, GPT, Gemini, DeepSeek, etc.).[/]"
)
self.console.print(
"[dim] Switch with: /model sonnet or /model gpt5[/]"
)
self.console.print()
def _preload_resumed_session(self) -> bool: