feat(gateway): add fast mode support to gateway chats

gateway/run.py

@@ -481,6 +481,7 @@ class GatewayRunner:
         self._prefill_messages = self._load_prefill_messages()
         self._ephemeral_system_prompt = self._load_ephemeral_system_prompt()
         self._reasoning_config = self._load_reasoning_config()
+        self._service_tier = self._load_service_tier()
         self._show_reasoning = self._load_show_reasoning()
         self._provider_routing = self._load_provider_routing()
         self._fallback_model = self._load_fallback_model()
@@ -776,6 +777,7 @@ class GatewayRunner:
 
     def _resolve_turn_agent_config(self, user_message: str, model: str, runtime_kwargs: dict) -> dict:
         from agent.smart_model_routing import resolve_turn_route
+        from hermes_cli.models import resolve_fast_mode_overrides
 
         primary = {
             "model": model,
@@ -787,7 +789,19 @@ class GatewayRunner:
             "args": list(runtime_kwargs.get("args") or []),
             "credential_pool": runtime_kwargs.get("credential_pool"),
         }
-        return resolve_turn_route(user_message, getattr(self, "_smart_model_routing", {}), primary)
+        route = resolve_turn_route(user_message, getattr(self, "_smart_model_routing", {}), primary)
+
+        service_tier = getattr(self, "_service_tier", None)
+        if not service_tier:
+            route["request_overrides"] = None
+            return route
+
+        try:
+            overrides = resolve_fast_mode_overrides(route.get("model"))
+        except Exception:
+            overrides = None
+        route["request_overrides"] = overrides
+        return route
 
     async def _handle_adapter_fatal_error(self, adapter: BasePlatformAdapter) -> None:
         """React to an adapter failure after startup.
@@ -939,6 +953,33 @@ class GatewayRunner:
             logger.warning("Unknown reasoning_effort '%s', using default (medium)", effort)
         return result
 
+    @staticmethod
+    def _load_service_tier() -> str | None:
+        """Load Priority Processing setting from config.yaml.
+
+        Reads agent.service_tier from config.yaml. Accepted values mirror the CLI:
+        "fast"/"priority"/"on" => "priority", while "normal"/"off" disables it.
+        Returns None when unset or unsupported.
+        """
+        raw = ""
+        try:
+            import yaml as _y
+            cfg_path = _hermes_home / "config.yaml"
+            if cfg_path.exists():
+                with open(cfg_path, encoding="utf-8") as _f:
+                    cfg = _y.safe_load(_f) or {}
+                raw = str(cfg.get("agent", {}).get("service_tier", "") or "").strip()
+        except Exception:
+            pass
+
+        value = raw.lower()
+        if not value or value in {"normal", "default", "standard", "off", "none"}:
+            return None
+        if value in {"fast", "priority", "on"}:
+            return "priority"
+        logger.warning("Unknown service_tier '%s', ignoring", raw)
+        return None
+
     @staticmethod
     def _load_show_reasoning() -> bool:
         """Load show_reasoning toggle from config.yaml display section."""
@@ -2088,6 +2129,9 @@ class GatewayRunner:
         if canonical == "reasoning":
             return await self._handle_reasoning_command(event)
 
+        if canonical == "fast":
+            return await self._handle_fast_command(event)
+
         if canonical == "verbose":
             return await self._handle_verbose_command(event)
 
@@ -4602,6 +4646,7 @@ class GatewayRunner:
             max_iterations = int(os.getenv("HERMES_MAX_ITERATIONS", "90"))
             reasoning_config = self._load_reasoning_config()
             self._reasoning_config = reasoning_config
+            self._service_tier = self._load_service_tier()
             turn_route = self._resolve_turn_agent_config(prompt, model, runtime_kwargs)
 
             def run_sync():
@@ -4613,6 +4658,8 @@ class GatewayRunner:
                     verbose_logging=False,
                     enabled_toolsets=enabled_toolsets,
                     reasoning_config=reasoning_config,
+                    service_tier=self._service_tier,
+                    request_overrides=turn_route.get("request_overrides"),
                     providers_allowed=pr.get("only"),
                     providers_ignored=pr.get("ignore"),
                     providers_order=pr.get("order"),
@@ -4762,6 +4809,7 @@ class GatewayRunner:
             model = _resolve_gateway_model(user_config)
             platform_key = _platform_config_key(source.platform)
             reasoning_config = self._load_reasoning_config()
+            self._service_tier = self._load_service_tier()
             turn_route = self._resolve_turn_agent_config(question, model, runtime_kwargs)
             pr = self._provider_routing
 
@@ -4788,6 +4836,8 @@ class GatewayRunner:
                     verbose_logging=False,
                     enabled_toolsets=[],
                     reasoning_config=reasoning_config,
+                    service_tier=self._service_tier,
+                    request_overrides=turn_route.get("request_overrides"),
                     providers_allowed=pr.get("only"),
                     providers_ignored=pr.get("ignore"),
                     providers_order=pr.get("order"),
@@ -4941,6 +4991,66 @@ class GatewayRunner:
         else:
             return f"🧠 ✓ Reasoning effort set to `{effort}` (this session only)"
 
+    async def _handle_fast_command(self, event: MessageEvent) -> str:
+        """Handle /fast — mirror the CLI Priority Processing toggle in gateway chats."""
+        import yaml
+        from hermes_cli.models import model_supports_fast_mode
+
+        args = event.get_command_args().strip().lower()
+        config_path = _hermes_home / "config.yaml"
+        self._service_tier = self._load_service_tier()
+
+        user_config = _load_gateway_config()
+        model = _resolve_gateway_model(user_config)
+        if not model_supports_fast_mode(model):
+            return "⚡ /fast is only available for OpenAI models that support Priority Processing."
+
+        def _save_config_key(key_path: str, value):
+            """Save a dot-separated key to config.yaml."""
+            try:
+                user_config = {}
+                if config_path.exists():
+                    with open(config_path, encoding="utf-8") as f:
+                        user_config = yaml.safe_load(f) or {}
+                keys = key_path.split(".")
+                current = user_config
+                for k in keys[:-1]:
+                    if k not in current or not isinstance(current[k], dict):
+                        current[k] = {}
+                    current = current[k]
+                current[keys[-1]] = value
+                atomic_yaml_write(config_path, user_config)
+                return True
+            except Exception as e:
+                logger.error("Failed to save config key %s: %s", key_path, e)
+                return False
+
+        if not args or args == "status":
+            status = "fast" if self._service_tier == "priority" else "normal"
+            return (
+                "⚡ Priority Processing\n\n"
+                f"Current mode: `{status}`\n\n"
+                "_Usage:_ `/fast <normal|fast|status>`"
+            )
+
+        if args in {"fast", "on"}:
+            self._service_tier = "priority"
+            saved_value = "fast"
+            label = "FAST"
+        elif args in {"normal", "off"}:
+            self._service_tier = None
+            saved_value = "normal"
+            label = "NORMAL"
+        else:
+            return (
+                f"⚠️ Unknown argument: `{args}`\n\n"
+                "**Valid options:** normal, fast, status"
+            )
+
+        if _save_config_key("agent.service_tier", saved_value):
+            return f"⚡ ✓ Priority Processing: **{label}** (saved to config)\n_(takes effect on next message)_"
+        return f"⚡ ✓ Priority Processing: **{label}** (this session only)"
+
     async def _handle_yolo_command(self, event: MessageEvent) -> str:
         """Handle /yolo — toggle dangerous command approval bypass for this session only."""
         from tools.approval import (
@@ -6771,6 +6881,7 @@ class GatewayRunner:
             pr = self._provider_routing
             reasoning_config = self._load_reasoning_config()
             self._reasoning_config = reasoning_config
+            self._service_tier = self._load_service_tier()
             # Set up streaming consumer if enabled
             _stream_consumer = None
             _stream_delta_cb = None
@@ -6833,6 +6944,8 @@ class GatewayRunner:
                     ephemeral_system_prompt=combined_ephemeral or None,
                     prefill_messages=self._prefill_messages or None,
                     reasoning_config=reasoning_config,
+                    service_tier=self._service_tier,
+                    request_overrides=turn_route.get("request_overrides"),
                     providers_allowed=pr.get("only"),
                     providers_ignored=pr.get("ignore"),
                     providers_order=pr.get("order"),
@@ -6857,6 +6970,8 @@ class GatewayRunner:
             agent.stream_delta_callback = _stream_delta_cb
             agent.status_callback = _status_callback_sync
             agent.reasoning_config = reasoning_config
+            agent.service_tier = self._service_tier
+            agent.request_overrides = turn_route.get("request_overrides")
 
             # Background review delivery — send "💾 Memory updated" etc. to user
             def _bg_review_send(message: str) -> None:

hermes_cli/commands.py

@@ -110,7 +110,7 @@ COMMAND_REGISTRY: list[CommandDef] = [
                args_hint="[level|show|hide]",
                subcommands=("none", "minimal", "low", "medium", "high", "xhigh", "show", "hide", "on", "off")),
     CommandDef("fast", "Toggle fast mode — OpenAI Priority Processing / Anthropic Fast Mode (Normal/Fast)", "Configuration",
-               cli_only=True, args_hint="[normal|fast|status]",
+               args_hint="[normal|fast|status]",
                subcommands=("normal", "fast", "status", "on", "off")),
     CommandDef("skin", "Show or change the display skin/theme", "Configuration",
                cli_only=True, args_hint="[name]"),

tests/gateway/test_fast_command.py

@@ -0,0 +1,190 @@
+"""Tests for gateway /fast support and Priority Processing routing."""
+
+import sys
+import threading
+import types
+from types import SimpleNamespace
+from unittest.mock import AsyncMock, patch
+
+import pytest
+import yaml
+
+import gateway.run as gateway_run
+from gateway.config import Platform
+from gateway.platforms.base import MessageEvent
+from gateway.session import SessionSource
+
+
+class _CapturingAgent:
+    last_init = None
+    last_run = None
+
+    def __init__(self, *args, **kwargs):
+        type(self).last_init = dict(kwargs)
+        self.tools = []
+
+    def run_conversation(self, user_message, conversation_history=None, task_id=None, persist_user_message=None):
+        type(self).last_run = {
+            "user_message": user_message,
+            "conversation_history": conversation_history,
+            "task_id": task_id,
+            "persist_user_message": persist_user_message,
+        }
+        return {
+            "final_response": "ok",
+            "messages": [],
+            "api_calls": 1,
+            "completed": True,
+        }
+
+
+def _install_fake_agent(monkeypatch):
+    fake_run_agent = types.ModuleType("run_agent")
+    fake_run_agent.AIAgent = _CapturingAgent
+    monkeypatch.setitem(sys.modules, "run_agent", fake_run_agent)
+
+
+def _make_runner():
+    runner = object.__new__(gateway_run.GatewayRunner)
+    runner.adapters = {}
+    runner._ephemeral_system_prompt = ""
+    runner._prefill_messages = []
+    runner._reasoning_config = None
+    runner._service_tier = None
+    runner._provider_routing = {}
+    runner._fallback_model = None
+    runner._smart_model_routing = {}
+    runner._running_agents = {}
+    runner._pending_model_notes = {}
+    runner._session_db = None
+    runner._agent_cache = {}
+    runner._agent_cache_lock = threading.Lock()
+    runner.hooks = SimpleNamespace(loaded_hooks=False)
+    runner.config = SimpleNamespace(streaming=None)
+    runner.session_store = SimpleNamespace(
+        get_or_create_session=lambda source: SimpleNamespace(session_id="session-1"),
+        load_transcript=lambda session_id: [],
+    )
+    runner._get_or_create_gateway_honcho = lambda session_key: (None, None)
+    runner._enrich_message_with_vision = AsyncMock(return_value="ENRICHED")
+    return runner
+
+
+def _make_source() -> SessionSource:
+    return SessionSource(
+        platform=Platform.TELEGRAM,
+        chat_id="12345",
+        chat_type="dm",
+        user_id="user-1",
+    )
+
+
+def _make_event(text: str) -> MessageEvent:
+    return MessageEvent(text=text, source=_make_source(), message_id="m1")
+
+
+def test_turn_route_injects_priority_processing_without_changing_runtime():
+    runner = _make_runner()
+    runner._service_tier = "priority"
+    runtime_kwargs = {
+        "api_key": "***",
+        "base_url": "https://openrouter.ai/api/v1",
+        "provider": "openrouter",
+        "api_mode": "chat_completions",
+        "command": None,
+        "args": [],
+        "credential_pool": None,
+    }
+
+    with patch("agent.smart_model_routing.resolve_turn_route", return_value={
+        "model": "gpt-5.4",
+        "runtime": dict(runtime_kwargs),
+        "label": None,
+        "signature": ("gpt-5.4", "openrouter", "https://openrouter.ai/api/v1", "chat_completions", None, ()),
+    }):
+        route = gateway_run.GatewayRunner._resolve_turn_agent_config(runner, "hi", "gpt-5.4", runtime_kwargs)
+
+    assert route["runtime"]["provider"] == "openrouter"
+    assert route["runtime"]["api_mode"] == "chat_completions"
+    assert route["request_overrides"] == {"service_tier": "priority"}
+
+
+def test_turn_route_skips_priority_processing_for_unsupported_models():
+    runner = _make_runner()
+    runner._service_tier = "priority"
+    runtime_kwargs = {
+        "api_key": "***",
+        "base_url": "https://openrouter.ai/api/v1",
+        "provider": "openrouter",
+        "api_mode": "chat_completions",
+        "command": None,
+        "args": [],
+        "credential_pool": None,
+    }
+
+    with patch("agent.smart_model_routing.resolve_turn_route", return_value={
+        "model": "gpt-5.3-codex",
+        "runtime": dict(runtime_kwargs),
+        "label": None,
+        "signature": ("gpt-5.3-codex", "openrouter", "https://openrouter.ai/api/v1", "chat_completions", None, ()),
+    }):
+        route = gateway_run.GatewayRunner._resolve_turn_agent_config(runner, "hi", "gpt-5.3-codex", runtime_kwargs)
+
+    assert route["request_overrides"] is None
+
+
+@pytest.mark.asyncio
+async def test_handle_fast_command_persists_config(monkeypatch, tmp_path):
+    runner = _make_runner()
+
+    monkeypatch.setattr(gateway_run, "_hermes_home", tmp_path)
+    monkeypatch.setattr(gateway_run, "_load_gateway_config", lambda: {})
+    monkeypatch.setattr(gateway_run, "_resolve_gateway_model", lambda config=None: "gpt-5.4")
+
+    response = await runner._handle_fast_command(_make_event("/fast fast"))
+
+    assert "FAST" in response
+    assert runner._service_tier == "priority"
+
+    saved = yaml.safe_load((tmp_path / "config.yaml").read_text(encoding="utf-8"))
+    assert saved["agent"]["service_tier"] == "fast"
+
+
+@pytest.mark.asyncio
+async def test_run_agent_passes_priority_processing_to_gateway_agent(monkeypatch, tmp_path):
+    _install_fake_agent(monkeypatch)
+    runner = _make_runner()
+
+    (tmp_path / "config.yaml").write_text("agent:\n  service_tier: fast\n", encoding="utf-8")
+    monkeypatch.setattr(gateway_run, "_hermes_home", tmp_path)
+    monkeypatch.setattr(gateway_run, "_env_path", tmp_path / ".env")
+    monkeypatch.setattr(gateway_run, "load_dotenv", lambda *args, **kwargs: None)
+    monkeypatch.setattr(gateway_run, "_load_gateway_config", lambda: {})
+    monkeypatch.setattr(gateway_run, "_resolve_gateway_model", lambda config=None: "gpt-5.4")
+    monkeypatch.setattr(
+        gateway_run,
+        "_resolve_runtime_agent_kwargs",
+        lambda: {
+            "provider": "openrouter",
+            "api_mode": "chat_completions",
+            "base_url": "https://openrouter.ai/api/v1",
+            "api_key": "***",
+        },
+    )
+
+    import hermes_cli.tools_config as tools_config
+    monkeypatch.setattr(tools_config, "_get_platform_tools", lambda user_config, platform_key: {"core"})
+
+    _CapturingAgent.last_init = None
+    result = await runner._run_agent(
+        message="hi",
+        context_prompt="",
+        history=[],
+        source=_make_source(),
+        session_id="session-1",
+        session_key="agent:main:telegram:dm:12345",
+    )
+
+    assert result["final_response"] == "ok"
+    assert _CapturingAgent.last_init["service_tier"] == "priority"
+    assert _CapturingAgent.last_init["request_overrides"] == {"service_tier": "priority"}