fix: accept max reasoning effort

batch_runner.py

@@ -1161,7 +1161,7 @@ def main(
         providers_order (str): Comma-separated list of OpenRouter providers to try in order (e.g. "anthropic,openai,google")
         provider_sort (str): Sort providers by "price", "throughput", or "latency" (OpenRouter only)
         max_tokens (int): Maximum tokens for model responses (optional, uses model default if not set)
-        reasoning_effort (str): OpenRouter reasoning effort level: "none", "minimal", "low", "medium", "high", "xhigh" (default: "medium")
+        reasoning_effort (str): OpenRouter reasoning effort level: "none", "minimal", "low", "medium", "high", "xhigh", "max" (default: "medium")
         reasoning_disabled (bool): Completely disable reasoning/thinking tokens (default: False)
         prefill_messages_file (str): Path to JSON file containing prefill messages (list of {role, content} dicts)
         max_samples (int): Only process the first N samples from the dataset (optional, processes all if not set)
@@ -1230,7 +1230,7 @@ def main(
         print("🧠 Reasoning: DISABLED (effort=none)")
     elif reasoning_effort:
         # Use specified effort level
-        valid_efforts = ["none", "minimal", "low", "medium", "high", "xhigh"]
+        valid_efforts = ["none", "minimal", "low", "medium", "high", "xhigh", "max"]
         if reasoning_effort not in valid_efforts:
             print(f"❌ Error: --reasoning_effort must be one of: {', '.join(valid_efforts)}")
             return

cli.py

@@ -6449,7 +6449,7 @@ def _handle_reasoning_command(self, cmd: str):
 
         Usage:
             /reasoning              Show current effort level and display state
-            /reasoning <level>      Set reasoning effort (none, minimal, low, medium, high, xhigh)
+            /reasoning <level>      Set reasoning effort (none, minimal, low, medium, high, xhigh, max)
             /reasoning show|on      Show model thinking/reasoning in output
             /reasoning hide|off     Hide model thinking/reasoning from output
         """
@@ -6493,7 +6493,7 @@ def _handle_reasoning_command(self, cmd: str):
         parsed = _parse_reasoning_config(arg)
         if parsed is None:
             _cprint(f"  {_DIM}(._.) Unknown argument: {arg}{_RST}")
-            _cprint(f"  {_DIM}Valid levels: none, minimal, low, medium, high, xhigh{_RST}")
+            _cprint(f"  {_DIM}Valid levels: none, minimal, low, medium, high, xhigh, max{_RST}")
             _cprint(f"  {_DIM}Display:      show, hide{_RST}")
             return
 

gateway/run.py

@@ -1189,7 +1189,7 @@ def _load_reasoning_config() -> dict | None:
         """Load reasoning effort from config.yaml.
 
         Reads agent.reasoning_effort from config.yaml. Valid: "none",
-        "minimal", "low", "medium", "high", "xhigh". Returns None to use
+        "minimal", "low", "medium", "high", "xhigh", "max". Returns None to use
         default (medium).
         """
         from hermes_constants import parse_reasoning_effort
@@ -6371,7 +6371,7 @@ async def _handle_reasoning_command(self, event: MessageEvent) -> str:
 
         Usage:
             /reasoning              Show current effort level and display state
-            /reasoning <level>      Set reasoning effort (none, minimal, low, medium, high, xhigh)
+            /reasoning <level>      Set reasoning effort (none, minimal, low, medium, high, xhigh, max)
             /reasoning show|on      Show model reasoning in responses
             /reasoning hide|off     Hide model reasoning from responses
         """
@@ -6438,12 +6438,12 @@ def _save_config_key(key_path: str, value):
         effort = args.strip()
         if effort == "none":
             parsed = {"enabled": False}
-        elif effort in ("minimal", "low", "medium", "high", "xhigh"):
+        elif effort in ("minimal", "low", "medium", "high", "xhigh", "max"):
             parsed = {"enabled": True, "effort": effort}
         else:
             return (
                 f"⚠️ Unknown argument: `{effort}`\n\n"
-                "**Valid levels:** none, minimal, low, medium, high, xhigh\n"
+                "**Valid levels:** none, minimal, low, medium, high, xhigh, max\n"
                 "**Display:** show, hide"
             )
 

hermes_cli/commands.py

@@ -119,7 +119,7 @@ class CommandDef:
                "Configuration"),
     CommandDef("reasoning", "Manage reasoning effort and display", "Configuration",
                args_hint="[level|show|hide]",
-               subcommands=("none", "minimal", "low", "medium", "high", "xhigh", "show", "hide", "on", "off")),
+               subcommands=("none", "minimal", "low", "medium", "high", "xhigh", "max", "show", "hide", "on", "off")),
     CommandDef("fast", "Toggle fast mode — OpenAI Priority Processing / Anthropic Fast Mode (Normal/Fast)", "Configuration",
                args_hint="[normal|fast|status]",
                subcommands=("normal", "fast", "status", "on", "off")),

hermes_cli/main.py

@@ -2872,7 +2872,7 @@ def _prompt_reasoning_effort_selection(efforts, current_effort=""):
             str(effort).strip().lower() for effort in efforts if str(effort).strip()
         )
     )
-    canonical_order = ("minimal", "low", "medium", "high", "xhigh")
+    canonical_order = ("minimal", "low", "medium", "high", "xhigh", "max")
     ordered = [effort for effort in canonical_order if effort in deduped]
     ordered.extend(effort for effort in deduped if effort not in canonical_order)
     if not ordered:

hermes_constants.py

@@ -138,13 +138,13 @@ def get_subprocess_home() -> str | None:
     return None
 
 
-VALID_REASONING_EFFORTS = ("minimal", "low", "medium", "high", "xhigh")
+VALID_REASONING_EFFORTS = ("minimal", "low", "medium", "high", "xhigh", "max")
 
 
 def parse_reasoning_effort(effort: str) -> dict | None:
     """Parse a reasoning effort level into a config dict.
 
-    Valid levels: "none", "minimal", "low", "medium", "high", "xhigh".
+    Valid levels: "none", "minimal", "low", "medium", "high", "xhigh", "max".
     Returns None when the input is empty or unrecognized (caller uses default).
     Returns {"enabled": False} for "none".
     Returns {"enabled": True, "effort": <level>} for valid effort levels.

skills/autonomous-ai-agents/hermes-agent/SKILL.md

@@ -250,7 +250,7 @@ Type these during an interactive chat session.
 /model [name]        Show or change model
 /provider            Show provider info
 /personality [name]  Set personality
-/reasoning [level]   Set reasoning (none|minimal|low|medium|high|xhigh|show|hide)
+/reasoning [level]   Set reasoning (none|minimal|low|medium|high|xhigh|max|show|hide)
 /verbose             Cycle: off → new → all → verbose
 /voice [on|off|tts]  Voice mode
 /yolo                Toggle approval bypass

tests/cli/test_reasoning_command.py

@@ -30,7 +30,7 @@ def test_none_disables(self):
         self.assertEqual(result, {"enabled": False})
 
     def test_valid_levels(self):
-        for level in ("low", "medium", "high", "xhigh", "minimal"):
+        for level in ("low", "medium", "high", "xhigh", "max", "minimal"):
             result = self._parse(level)
             self.assertIsNotNone(result)
             self.assertTrue(result.get("enabled"))

tests/gateway/test_reasoning_command.py

@@ -118,6 +118,25 @@ async def test_handle_reasoning_command_updates_config_and_cache(self, tmp_path,
         assert runner._reasoning_config == {"enabled": True, "effort": "low"}
         assert "takes effect on next message" in result
 
+    @pytest.mark.asyncio
+    async def test_handle_reasoning_command_accepts_max(self, tmp_path, monkeypatch):
+        hermes_home = tmp_path / "hermes"
+        hermes_home.mkdir()
+        config_path = hermes_home / "config.yaml"
+        config_path.write_text("agent:\n  reasoning_effort: medium\n", encoding="utf-8")
+
+        monkeypatch.setattr(gateway_run, "_hermes_home", hermes_home)
+
+        runner = _make_runner()
+        runner._reasoning_config = {"enabled": True, "effort": "medium"}
+
+        result = await runner._handle_reasoning_command(_make_event("/reasoning max"))
+
+        saved = yaml.safe_load(config_path.read_text(encoding="utf-8"))
+        assert saved["agent"]["reasoning_effort"] == "max"
+        assert runner._reasoning_config == {"enabled": True, "effort": "max"}
+        assert "takes effect on next message" in result
+
     def test_run_agent_reloads_reasoning_config_per_message(self, tmp_path, monkeypatch):
         hermes_home = tmp_path / "hermes"
         hermes_home.mkdir()

tests/hermes_cli/test_reasoning_effort_menu.py

@@ -32,3 +32,21 @@ def test_reasoning_menu_orders_minimal_before_low(monkeypatch):
         "  medium  ← currently in use",
         "  high",
     ]
+
+
+def test_reasoning_menu_places_max_after_xhigh(monkeypatch):
+    fake_module = types.SimpleNamespace(TerminalMenu=_FakeTerminalMenu)
+    monkeypatch.setitem(sys.modules, "simple_term_menu", fake_module)
+
+    selected = _prompt_reasoning_effort_selection(
+        ["max", "medium", "xhigh", "high"],
+        current_effort="max",
+    )
+
+    assert selected == "max"
+    assert _FakeTerminalMenu.last_choices[:4] == [
+        "  medium",
+        "  high",
+        "  xhigh",
+        "  max  ← currently in use",
+    ]

website/docs/user-guide/configuration.md

@@ -795,7 +795,7 @@ Control how much "thinking" the model does before responding:
 
 ```yaml
 agent:
-  reasoning_effort: ""   # empty = medium (default). Options: none, minimal, low, medium, high, xhigh (max)
+  reasoning_effort: ""   # empty = medium (default). Options: none, minimal, low, medium, high, xhigh, max
 ```
 
 When unset (default), reasoning effort defaults to "medium" — a balanced level that works well for most tasks. Setting a value overrides it — higher reasoning effort gives better results on complex tasks at the cost of more tokens and latency.

website/docs/user-guide/features/batch-processing.md

@@ -79,7 +79,7 @@ Entries can optionally include:
 
 | Parameter | Description |
 |-----------|-------------|
-| `--reasoning_effort` | Effort level: `none`, `minimal`, `low`, `medium`, `high`, `xhigh` |
+| `--reasoning_effort` | Effort level: `none`, `minimal`, `low`, `medium`, `high`, `xhigh`, `max` |
 | `--reasoning_disabled` | Completely disable reasoning/thinking tokens |
 
 ### Advanced Options