NousResearch/hermes-agent · critical · RuntimeError

No LLM provider configured. Run `hermes model` to select a p

Error message

No LLM provider configured. Run `hermes model` to select a provider, or run `hermes setup` for first-time configuration.

What it means

Sibling of the explicit-provider failure: raised at init when NO provider is configured at all and no fallback client was activated (agent._fallback_activated is False). Hermes deliberately rejects startup rather than silently defaulting to an unintended provider.

Source

Thrown at agent/agent_init.py:1359

                                client_kwargs["timeout"] = _provider_timeout
                            _fb_headers = getattr(_fb_client, "_custom_headers", None)
                            if not _fb_headers:
                                _fb_headers = getattr(_fb_client, "default_headers", None)
                            if not _fb_headers:
                                _fb_headers = getattr(_fb_client, "_default_headers", None)
                            if _fb_headers:
                                client_kwargs["default_headers"] = dict(_fb_headers)
                            _fb_resolved = True
                            break
                    if not _fb_resolved:
                        raise RuntimeError(
                            f"Provider '{_explicit}' is set in config.yaml but no API key "
                            f"was found. Set the {_env_hint} environment "
                            f"variable, or switch to a different provider with `hermes model`."
                        )
                if not getattr(agent, "_fallback_activated", False):
                    # No provider configured — reject with a clear message.
                    raise RuntimeError(
                        "No LLM provider configured. Run `hermes model` to "
                        "select a provider, or run `hermes setup` for first-time "
                        "configuration."
                    )
        
        agent._client_kwargs = client_kwargs  # stored for rebuilding after interrupt

        # Enable fine-grained tool streaming for Claude on OpenRouter.
        # Without this, Anthropic buffers the entire tool call and goes
        # silent for minutes while thinking — OpenRouter's upstream proxy
        # times out during the silence.  The beta header makes Anthropic
        # stream tool call arguments token-by-token, keeping the
        # connection alive.
        _effective_base = str(client_kwargs.get("base_url", "")).lower()
        if base_url_host_matches(_effective_base, "openrouter.ai") and "claude" in (agent.model or "").lower():
            headers = client_kwargs.get("default_headers") or {}
            existing_beta = headers.get("x-anthropic-beta", "")
            _FINE_GRAINED = "fine-grained-tool-streaming-2025-05-14"

View on GitHub (pinned to c896c09c42)

Solutions

Run `hermes model` to select a provider and model
Run `hermes setup` for guided first-time setup (provider + key + model)
Restore ~/.hermes/config.yaml from backup if it was accidentally removed
Confirm the active profile's HERMES_HOME actually contains a config.yaml with a model section

Defensive patterns

Strategy: validation

Validate before calling

from pathlib import Path
from hermes_constants import get_hermes_home
import yaml

def provider_configured() -> bool:
    cfg_path = get_hermes_home() / "config.yaml"
    if not cfg_path.exists():
        return False
    model = (yaml.safe_load(cfg_path.read_text()) or {}).get("model") or {}
    return bool(model.get("provider") or model.get("base_url"))

if not provider_configured():
    run_first_time_setup()  # hermes setup / hermes model

Try / catch

try:
    agent = AIAgent(...)
except RuntimeError as e:
    if "No LLM provider configured" in str(e):
        launch_setup_wizard()
    else:
        raise

Prevention

Run `hermes setup` once on fresh installs before any scripted agent usage
When creating profiles, clone or seed config.yaml so a model.provider is always present
Smoke-test deployments with a trivial `hermes chat 'ping'` to catch missing-provider early

When it happens

Trigger: Fresh HERMES_HOME with empty/minimal config.yaml (no model.provider, no custom endpoint), no provider env vars discoverable, and no credential-pool entry that would activate a fallback — every resolution tier returned nothing.

Common situations: First run before `hermes setup`; config.yaml deleted, reset, or lost; a profile created without copying config; .env wiped so no keys are discoverable by any tier.

Related errors

AI-assisted analysis of NousResearch/hermes-agent@c896c09c42 (2026-08-14). Data as JSON: /api/errors/d65bdaa8e94f7992. Report an issue: GitHub.