feat: add reasoning effort configuration for agent

- Introduced a new configuration option for reasoning effort in the CLI, allowing users to specify the level of reasoning the agent should perform before responding.
- Updated the CLI and agent initialization to incorporate the reasoning configuration, enhancing the agent's responsiveness and adaptability.
- Implemented logic to load reasoning effort from environment variables and configuration files, providing flexibility in agent behavior.
- Enhanced the documentation in the example configuration file to clarify the new reasoning effort options available.

cli-config.yaml.example

@@ -210,9 +210,10 @@ agent:
   # Enable verbose logging
   verbose: false
   
-  # Custom system prompt (personality, instructions, etc.)
+  # Reasoning effort level (OpenRouter and Nous Portal)
-  # Leave empty or remove to use default agent behavior
+  # Controls how much "thinking" the model does before responding.
-  system_prompt: ""
+  # Options: "xhigh" (max), "high", "medium", "low", "minimal", "none" (disable)
+  reasoning_effort: "xhigh"
   
   # Predefined personalities (use with /personality command)
   personalities:

cli.py

@@ -90,6 +90,24 @@ def _load_prefill_messages(file_path: str) -> List[Dict[str, Any]]:
         return []
 
 
+def _parse_reasoning_config(effort: str) -> dict | None:
+    """Parse a reasoning effort level into an OpenRouter reasoning config dict.
+    
+    Valid levels: "xhigh", "high", "medium", "low", "minimal", "none".
+    Returns None to use the default (xhigh), or a config dict to override.
+    """
+    if not effort or not effort.strip():
+        return None
+    effort = effort.strip().lower()
+    if effort == "none":
+        return {"enabled": False}
+    valid = ("xhigh", "high", "medium", "low", "minimal")
+    if effort in valid:
+        return {"enabled": True, "effort": effort}
+    logger.warning("Unknown reasoning_effort '%s', using default (xhigh)", effort)
+    return None
+
+
 def load_cli_config() -> Dict[str, Any]:
     """
     Load CLI configuration from config files.
@@ -146,6 +164,7 @@ def load_cli_config() -> Dict[str, Any]:
             "verbose": False,
             "system_prompt": "",
             "prefill_messages_file": "",
+            "reasoning_effort": "",
             "personalities": {
                 "helpful": "You are a helpful, friendly AI assistant.",
                 "concise": "You are a concise assistant. Keep responses brief and to the point.",
@@ -795,6 +814,11 @@ class HermesCLI:
             CLI_CONFIG["agent"].get("prefill_messages_file", "")
         )
         
+        # Reasoning config (OpenRouter reasoning effort level)
+        self.reasoning_config = _parse_reasoning_config(
+            CLI_CONFIG["agent"].get("reasoning_effort", "")
+        )
+        
         # Agent will be initialized on first use
         self.agent: Optional[AIAgent] = None
         self._app = None  # prompt_toolkit Application (set in run())
@@ -889,6 +913,7 @@ class HermesCLI:
                 quiet_mode=True,
                 ephemeral_system_prompt=self.system_prompt if self.system_prompt else None,
                 prefill_messages=self.prefill_messages or None,
+                reasoning_config=self.reasoning_config,
                 session_id=self.session_id,
                 platform="cli",
                 session_db=self._session_db,

gateway/run.py

@@ -96,6 +96,7 @@ class GatewayRunner:
         # Both are injected at API-call time only and never persisted.
         self._prefill_messages = self._load_prefill_messages()
         self._ephemeral_system_prompt = self._load_ephemeral_system_prompt()
+        self._reasoning_config = self._load_reasoning_config()
 
         # Wire process registry into session store for reset protection
         from tools.process_registry import process_registry
@@ -184,6 +185,36 @@ class GatewayRunner:
             pass
         return ""
 
+    @staticmethod
+    def _load_reasoning_config() -> dict | None:
+        """Load reasoning effort from config or env var.
+        
+        Checks HERMES_REASONING_EFFORT env var first, then agent.reasoning_effort
+        in config.yaml. Valid: "xhigh", "high", "medium", "low", "minimal", "none".
+        Returns None to use default (xhigh).
+        """
+        effort = os.getenv("HERMES_REASONING_EFFORT", "")
+        if not effort:
+            try:
+                import yaml as _y
+                cfg_path = Path.home() / ".hermes" / "config.yaml"
+                if cfg_path.exists():
+                    with open(cfg_path) as _f:
+                        cfg = _y.safe_load(_f) or {}
+                    effort = str(cfg.get("agent", {}).get("reasoning_effort", "") or "").strip()
+            except Exception:
+                pass
+        if not effort:
+            return None
+        effort = effort.lower().strip()
+        if effort == "none":
+            return {"enabled": False}
+        valid = ("xhigh", "high", "medium", "low", "minimal")
+        if effort in valid:
+            return {"enabled": True, "effort": effort}
+        logger.warning("Unknown reasoning_effort '%s', using default (xhigh)", effort)
+        return None
+
     async def start(self) -> bool:
         """
         Start the gateway and all configured platform adapters.
@@ -1352,6 +1383,7 @@ class GatewayRunner:
                 enabled_toolsets=enabled_toolsets,
                 ephemeral_system_prompt=combined_ephemeral or None,
                 prefill_messages=self._prefill_messages or None,
+                reasoning_config=self._reasoning_config,
                 session_id=session_id,
                 tool_progress_callback=progress_callback if tool_progress_enabled else None,
                 platform=platform_key,

run_agent.py

@@ -1152,7 +1152,11 @@ class AIAgent:
         if provider_preferences:
             extra_body["provider"] = provider_preferences
 
-        if "openrouter" in self.base_url.lower():
+        _supports_reasoning = (
+            "openrouter" in self.base_url.lower()
+            or "nousresearch" in self.base_url.lower()
+        )
+        if _supports_reasoning:
             if self.reasoning_config is not None:
                 extra_body["reasoning"] = self.reasoning_config
             else:
@@ -1574,7 +1578,11 @@ class AIAgent:
                     api_messages.insert(sys_offset + idx, pfm.copy())
 
             summary_extra_body = {}
-            if "openrouter" in self.base_url.lower():
+            _supports_reasoning = (
+                "openrouter" in self.base_url.lower()
+                or "nousresearch" in self.base_url.lower()
+            )
+            if _supports_reasoning:
                 if self.reasoning_config is not None:
                     summary_extra_body["reasoning"] = self.reasoning_config
                 else: