feat: add Codex fast mode toggle (/fast command)
Add /fast slash command to toggle OpenAI Codex service_tier between
normal and priority ('fast') inference. Only exposed for models
registered in _FAST_MODE_BACKEND_CONFIG (currently gpt-5.4).
- Registry-based backend config for extensibility
- Dynamic command visibility (hidden from help/autocomplete for
non-supported models) via command_filter on SlashCommandCompleter
- service_tier flows through request_overrides from route resolution
- Omit max_output_tokens for Codex backend (rejects it)
- Persists to config.yaml under agent.service_tier
Salvage cleanup: removed simple_term_menu/input() menu (banned),
bare /fast now shows status like /reasoning. Removed redundant
override resolution in _build_api_kwargs — single source of truth
via request_overrides from route.
Co-authored-by: Hermes Agent <hermes@nousresearch.com>
This commit is contained in:
g-guthrie
217
tests/cli/test_fast_command.py
Normal file
217
tests/cli/test_fast_command.py
Normal file
@@ -0,0 +1,217 @@
|
||||
"""Tests for the /fast CLI command and service-tier config handling."""
|
||||
|
||||
import unittest
|
||||
from types import SimpleNamespace
|
||||
from unittest.mock import MagicMock, patch
|
||||
|
||||
|
||||
def _import_cli():
|
||||
import hermes_cli.config as config_mod
|
||||
|
||||
if not hasattr(config_mod, "save_env_value_secure"):
|
||||
config_mod.save_env_value_secure = lambda key, value: {
|
||||
"success": True,
|
||||
"stored_as": key,
|
||||
"validated": False,
|
||||
}
|
||||
|
||||
import cli as cli_mod
|
||||
|
||||
return cli_mod
|
||||
|
||||
|
||||
class TestParseServiceTierConfig(unittest.TestCase):
|
||||
def _parse(self, raw):
|
||||
cli_mod = _import_cli()
|
||||
return cli_mod._parse_service_tier_config(raw)
|
||||
|
||||
def test_fast_maps_to_priority(self):
|
||||
self.assertEqual(self._parse("fast"), "priority")
|
||||
self.assertEqual(self._parse("priority"), "priority")
|
||||
|
||||
def test_normal_disables_service_tier(self):
|
||||
self.assertIsNone(self._parse("normal"))
|
||||
self.assertIsNone(self._parse("off"))
|
||||
self.assertIsNone(self._parse(""))
|
||||
|
||||
|
||||
class TestHandleFastCommand(unittest.TestCase):
|
||||
def _make_cli(self, service_tier=None):
|
||||
return SimpleNamespace(
|
||||
service_tier=service_tier,
|
||||
provider="openai-codex",
|
||||
requested_provider="openai-codex",
|
||||
model="gpt-5.4",
|
||||
_fast_command_available=lambda: True,
|
||||
agent=MagicMock(),
|
||||
)
|
||||
|
||||
def test_no_args_shows_status(self):
|
||||
cli_mod = _import_cli()
|
||||
stub = self._make_cli(service_tier=None)
|
||||
with (
|
||||
patch.object(cli_mod, "_cprint") as mock_cprint,
|
||||
patch.object(cli_mod, "save_config_value") as mock_save,
|
||||
):
|
||||
cli_mod.HermesCLI._handle_fast_command(stub, "/fast")
|
||||
|
||||
# Bare /fast shows status, does not change config
|
||||
mock_save.assert_not_called()
|
||||
# Should have printed the status line
|
||||
printed = " ".join(str(c) for c in mock_cprint.call_args_list)
|
||||
self.assertIn("normal", printed)
|
||||
|
||||
def test_no_args_shows_fast_when_enabled(self):
|
||||
cli_mod = _import_cli()
|
||||
stub = self._make_cli(service_tier="priority")
|
||||
with (
|
||||
patch.object(cli_mod, "_cprint") as mock_cprint,
|
||||
patch.object(cli_mod, "save_config_value") as mock_save,
|
||||
):
|
||||
cli_mod.HermesCLI._handle_fast_command(stub, "/fast")
|
||||
|
||||
mock_save.assert_not_called()
|
||||
printed = " ".join(str(c) for c in mock_cprint.call_args_list)
|
||||
self.assertIn("fast", printed)
|
||||
|
||||
def test_normal_argument_clears_service_tier(self):
|
||||
cli_mod = _import_cli()
|
||||
stub = self._make_cli(service_tier="priority")
|
||||
with (
|
||||
patch.object(cli_mod, "_cprint"),
|
||||
patch.object(cli_mod, "save_config_value", return_value=True) as mock_save,
|
||||
):
|
||||
cli_mod.HermesCLI._handle_fast_command(stub, "/fast normal")
|
||||
|
||||
mock_save.assert_called_once_with("agent.service_tier", "normal")
|
||||
self.assertIsNone(stub.service_tier)
|
||||
self.assertIsNone(stub.agent)
|
||||
|
||||
def test_unsupported_model_does_not_expose_fast(self):
|
||||
cli_mod = _import_cli()
|
||||
stub = SimpleNamespace(
|
||||
service_tier=None,
|
||||
provider="openai-codex",
|
||||
requested_provider="openai-codex",
|
||||
model="gpt-5.3-codex",
|
||||
_fast_command_available=lambda: False,
|
||||
agent=MagicMock(),
|
||||
)
|
||||
|
||||
with (
|
||||
patch.object(cli_mod, "_cprint") as mock_cprint,
|
||||
patch.object(cli_mod, "save_config_value") as mock_save,
|
||||
):
|
||||
cli_mod.HermesCLI._handle_fast_command(stub, "/fast")
|
||||
|
||||
mock_save.assert_not_called()
|
||||
self.assertTrue(mock_cprint.called)
|
||||
|
||||
|
||||
class TestFastModeRegistry(unittest.TestCase):
|
||||
def test_only_gpt_5_4_is_enabled_for_codex(self):
|
||||
from hermes_cli.models import fast_mode_backend_config
|
||||
|
||||
assert fast_mode_backend_config("gpt-5.4") == {
|
||||
"provider": "openai-codex",
|
||||
"request_overrides": {"service_tier": "priority"},
|
||||
}
|
||||
assert fast_mode_backend_config("gpt-5.3-codex") is None
|
||||
|
||||
|
||||
class TestFastModeRouting(unittest.TestCase):
|
||||
def test_fast_command_exposed_for_model_even_when_provider_is_auto(self):
|
||||
cli_mod = _import_cli()
|
||||
stub = SimpleNamespace(provider="auto", requested_provider="auto", model="gpt-5.4", agent=None)
|
||||
|
||||
assert cli_mod.HermesCLI._fast_command_available(stub) is True
|
||||
|
||||
def test_turn_route_switches_to_model_backend_when_fast_enabled(self):
|
||||
cli_mod = _import_cli()
|
||||
stub = SimpleNamespace(
|
||||
model="gpt-5.4",
|
||||
api_key="primary-key",
|
||||
base_url="https://openrouter.ai/api/v1",
|
||||
provider="openrouter",
|
||||
api_mode="chat_completions",
|
||||
acp_command=None,
|
||||
acp_args=[],
|
||||
_credential_pool=None,
|
||||
_smart_model_routing={},
|
||||
service_tier="priority",
|
||||
)
|
||||
|
||||
with (
|
||||
patch("agent.smart_model_routing.resolve_turn_route", return_value={
|
||||
"model": "gpt-5.4",
|
||||
"runtime": {
|
||||
"api_key": "primary-key",
|
||||
"base_url": "https://openrouter.ai/api/v1",
|
||||
"provider": "openrouter",
|
||||
"api_mode": "chat_completions",
|
||||
"command": None,
|
||||
"args": [],
|
||||
"credential_pool": None,
|
||||
},
|
||||
"label": None,
|
||||
"signature": ("gpt-5.4", "openrouter", "https://openrouter.ai/api/v1", "chat_completions", None, ()),
|
||||
}),
|
||||
patch("hermes_cli.runtime_provider.resolve_runtime_provider", return_value={
|
||||
"provider": "openai-codex",
|
||||
"api_mode": "codex_responses",
|
||||
"base_url": "https://chatgpt.com/backend-api/codex",
|
||||
"api_key": "codex-key",
|
||||
"command": None,
|
||||
"args": [],
|
||||
"credential_pool": None,
|
||||
}),
|
||||
):
|
||||
route = cli_mod.HermesCLI._resolve_turn_agent_config(stub, "hi")
|
||||
|
||||
assert route["runtime"]["provider"] == "openai-codex"
|
||||
assert route["runtime"]["api_mode"] == "codex_responses"
|
||||
assert route["request_overrides"] == {"service_tier": "priority"}
|
||||
|
||||
def test_turn_route_keeps_primary_runtime_when_model_has_no_fast_backend(self):
|
||||
cli_mod = _import_cli()
|
||||
stub = SimpleNamespace(
|
||||
model="gpt-5.3-codex",
|
||||
api_key="primary-key",
|
||||
base_url="https://openrouter.ai/api/v1",
|
||||
provider="openrouter",
|
||||
api_mode="chat_completions",
|
||||
acp_command=None,
|
||||
acp_args=[],
|
||||
_credential_pool=None,
|
||||
_smart_model_routing={},
|
||||
service_tier="priority",
|
||||
)
|
||||
|
||||
primary_route = {
|
||||
"model": "gpt-5.3-codex",
|
||||
"runtime": {
|
||||
"api_key": "primary-key",
|
||||
"base_url": "https://openrouter.ai/api/v1",
|
||||
"provider": "openrouter",
|
||||
"api_mode": "chat_completions",
|
||||
"command": None,
|
||||
"args": [],
|
||||
"credential_pool": None,
|
||||
},
|
||||
"label": None,
|
||||
"signature": ("gpt-5.3-codex", "openrouter", "https://openrouter.ai/api/v1", "chat_completions", None, ()),
|
||||
}
|
||||
with patch("agent.smart_model_routing.resolve_turn_route", return_value=primary_route):
|
||||
route = cli_mod.HermesCLI._resolve_turn_agent_config(stub, "hi")
|
||||
|
||||
assert route["runtime"]["provider"] == "openrouter"
|
||||
assert route.get("request_overrides") is None
|
||||
|
||||
|
||||
class TestConfigDefault(unittest.TestCase):
|
||||
def test_default_config_has_service_tier(self):
|
||||
from hermes_cli.config import DEFAULT_CONFIG
|
||||
|
||||
agent = DEFAULT_CONFIG.get("agent", {})
|
||||
self.assertIn("service_tier", agent)
|
||||
self.assertEqual(agent["service_tier"], "")
|
||||
@@ -446,6 +446,13 @@ class TestSubcommands:
|
||||
assert "show" in subs
|
||||
assert "hide" in subs
|
||||
|
||||
def test_fast_has_subcommands(self):
|
||||
assert "/fast" in SUBCOMMANDS
|
||||
subs = SUBCOMMANDS["/fast"]
|
||||
assert "fast" in subs
|
||||
assert "normal" in subs
|
||||
assert "status" in subs
|
||||
|
||||
def test_voice_has_subcommands(self):
|
||||
assert "/voice" in SUBCOMMANDS
|
||||
assert "on" in SUBCOMMANDS["/voice"]
|
||||
@@ -474,6 +481,20 @@ class TestSubcommandCompletion:
|
||||
assert "high" in texts
|
||||
assert "show" in texts
|
||||
|
||||
def test_fast_subcommand_completion_after_space(self):
|
||||
completions = _completions(SlashCommandCompleter(), "/fast ")
|
||||
texts = {c.text for c in completions}
|
||||
assert "fast" in texts
|
||||
assert "normal" in texts
|
||||
|
||||
def test_fast_command_filtered_out_when_unavailable(self):
|
||||
completions = _completions(
|
||||
SlashCommandCompleter(command_filter=lambda cmd: cmd != "/fast"),
|
||||
"/fa",
|
||||
)
|
||||
texts = {c.text for c in completions}
|
||||
assert "fast" not in texts
|
||||
|
||||
def test_subcommand_prefix_filters(self):
|
||||
"""Typing '/reasoning sh' should only show 'show'."""
|
||||
completions = _completions(SlashCommandCompleter(), "/reasoning sh")
|
||||
@@ -527,6 +548,13 @@ class TestGhostText:
|
||||
"""/reasoning sh → 'ow'"""
|
||||
assert _suggestion("/reasoning sh") == "ow"
|
||||
|
||||
def test_fast_subcommand_suggestion(self):
|
||||
assert _suggestion("/fast f") == "ast"
|
||||
|
||||
def test_fast_subcommand_suggestion_hidden_when_filtered(self):
|
||||
completer = SlashCommandCompleter(command_filter=lambda cmd: cmd != "/fast")
|
||||
assert _suggestion("/fa", completer=completer) is None
|
||||
|
||||
def test_no_suggestion_for_non_slash(self):
|
||||
assert _suggestion("hello") is None
|
||||
|
||||
|
||||
@@ -356,6 +356,25 @@ class TestBuildApiKwargsCodex:
|
||||
assert "reasoning" in kwargs
|
||||
assert kwargs["reasoning"]["effort"] == "medium"
|
||||
|
||||
def test_includes_service_tier_via_request_overrides(self, monkeypatch):
|
||||
agent = _make_agent(monkeypatch, "openai-codex", api_mode="codex_responses",
|
||||
base_url="https://chatgpt.com/backend-api/codex")
|
||||
agent.model = "gpt-5.4"
|
||||
agent.service_tier = "priority"
|
||||
agent.request_overrides = {"service_tier": "priority"}
|
||||
messages = [{"role": "user", "content": "hi"}]
|
||||
kwargs = agent._build_api_kwargs(messages)
|
||||
assert kwargs["service_tier"] == "priority"
|
||||
|
||||
def test_omits_max_output_tokens_for_codex_backend(self, monkeypatch):
|
||||
agent = _make_agent(monkeypatch, "openai-codex", api_mode="codex_responses",
|
||||
base_url="https://chatgpt.com/backend-api/codex")
|
||||
agent.model = "gpt-5.4"
|
||||
agent.max_tokens = 20
|
||||
messages = [{"role": "user", "content": "hi"}]
|
||||
kwargs = agent._build_api_kwargs(messages)
|
||||
assert "max_output_tokens" not in kwargs
|
||||
|
||||
def test_includes_encrypted_content_in_include(self, monkeypatch):
|
||||
agent = _make_agent(monkeypatch, "openai-codex", api_mode="codex_responses",
|
||||
base_url="https://chatgpt.com/backend-api/codex")
|
||||
|
||||
@@ -648,6 +648,15 @@ def test_preflight_codex_api_kwargs_allows_reasoning_and_temperature(monkeypatch
|
||||
assert result["max_output_tokens"] == 4096
|
||||
|
||||
|
||||
def test_preflight_codex_api_kwargs_allows_service_tier(monkeypatch):
|
||||
agent = _build_agent(monkeypatch)
|
||||
kwargs = _codex_request_kwargs()
|
||||
kwargs["service_tier"] = "priority"
|
||||
|
||||
result = agent._preflight_codex_api_kwargs(kwargs)
|
||||
assert result["service_tier"] == "priority"
|
||||
|
||||
|
||||
def test_run_conversation_codex_replay_payload_keeps_call_id(monkeypatch):
|
||||
agent = _build_agent(monkeypatch)
|
||||
responses = [_codex_tool_call_response(), _codex_message_response("done")]
|
||||
|
||||
Reference in New Issue
Block a user