refactor: trimmed some prompts

style: version tool output now mentions the default model in use

tests/test_auto_mode.py

@@ -93,7 +93,7 @@ class TestAutoMode:
             # Model field should have detailed descriptions
             model_schema = schema["properties"]["model"]
             assert "enum" not in model_schema
-            assert "auto model selection" in model_schema["description"].lower()
+            assert "auto mode" in model_schema["description"].lower()
             assert "listmodels" in model_schema["description"]
 
         finally:
@@ -289,7 +289,7 @@ class TestAutoMode:
             schema = tool.get_model_field_schema()
             assert "enum" not in schema
             assert schema["type"] == "string"
-            assert "auto model selection" in schema["description"]
+            assert "auto mode" in schema["description"].lower()
             assert "listmodels" in schema["description"]
 
             # Test normal mode

tests/test_chat_simple.py

@@ -87,7 +87,7 @@ class TestChatTool:
         # Description should route callers to listmodels, regardless of mode
         assert "listmodels" in schema["description"]
         if self.tool.is_effective_auto_mode():
-            assert "auto model selection" in schema["description"]
+            assert "auto mode" in schema["description"].lower()
         else:
             import config
 

tools/chat.py

@@ -25,10 +25,8 @@ CHAT_FIELD_DESCRIPTIONS = {
         "Your question or idea for collaborative thinking. Provide detailed context, including your goal, what you've tried, and any specific challenges. "
         "CRITICAL: To discuss code, use 'files' parameter instead of pasting code blocks here."
     ),
-    "files": "Always pass absolute full-paths (do NOT shorten) to existing files / folders containing code being discussed.",
+    "files": "absolute file or folder paths for code context (do NOT shorten).",
-    "images": (
+    "images": "Optional absolute image paths or base64 for visual context when helpful.",
-        "Optional images for visual context (must be FULL absolute paths to real files / folders - DO NOT SHORTEN - OR these can be bas64 data)"
-    ),
 }
 
 

tools/shared/base_models.py

@@ -20,24 +20,16 @@ logger = logging.getLogger(__name__)
 
 # Shared field descriptions to avoid duplication
 COMMON_FIELD_DESCRIPTIONS = {
-    "model": (
+    "model": "Model to run. Supply a name if requested by the user or stay in auto mode. When in auto mode, use `listmodels` tool for model discovery.",
-        "Model to use. See tool's input schema for available models if required. Use 'auto' select the best model for the task."
+    "temperature": "0 = deterministic · 1 = creative.",
-    ),
+    "thinking_mode": "Reasoning depth: minimal, low, medium, high, or max.",
-    "temperature": ("Lower values: deterministic; higher: creative."),
-    "thinking_mode": (
-        "Thinking depth: minimal (0.5%), low (8%), medium (33%), high (67%), "
-        "max (100% of model max). Higher modes: deeper reasoning but slower."
-    ),
     "continuation_id": (
         "Unique thread continuation ID for multi-turn conversations. Works across different tools. "
-        "ALWAYS reuse last continuation_id you were provided as-is when re-communicating with Zen MCP, "
+        "ALWAYS reuse the last continuation_id you were given—this preserves full conversation context, "
-        "unless user provides different ID. When supplied, your complete conversation history is available, so focus on new insights."
+        "files, and findings so the agent can resume seamlessly."
     ),
-    "images": (
+    "images": "Optional absolute image paths or base64 blobs for visual context.",
-        "Optional images for visual context. MUST be absolute paths or base64. "
+    "files": "Optional absolute file or folder paths (do not shorten).",
-        "Use when user mentions images. Describe image contents. "
-    ),
-    "files": ("Optional files for context (FULL absolute paths to real files/folders - DO NOT SHORTEN)"),
 }
 
 # Workflow-specific field descriptions

tools/version.py

@@ -193,6 +193,35 @@ class VersionTool(BaseTool):
         output_lines.append(f"**Last Updated**: {__updated__}")
         output_lines.append(f"**Author**: {__author__}")
 
+        model_selection_metadata = {"mode": "unknown", "default_model": None}
+        model_selection_display = "Model selection status unavailable"
+
+        # Model selection configuration
+        try:
+            from config import DEFAULT_MODEL
+            from tools.shared.base_tool import BaseTool
+
+            auto_mode = BaseTool.is_effective_auto_mode(self)
+            if auto_mode:
+                output_lines.append(
+                    "**Model Selection**: Auto model selection mode (call `listmodels` to inspect options)"
+                )
+                model_selection_metadata = {"mode": "auto", "default_model": DEFAULT_MODEL}
+                model_selection_display = "Auto model selection (use `listmodels` for options)"
+            else:
+                output_lines.append(f"**Model Selection**: Default model set to `{DEFAULT_MODEL}`")
+                model_selection_metadata = {"mode": "default", "default_model": DEFAULT_MODEL}
+                model_selection_display = f"Default model: `{DEFAULT_MODEL}`"
+        except Exception as exc:
+            logger.debug(f"Could not determine model selection mode: {exc}")
+
+        output_lines.append("")
+        output_lines.append("## Quick Summary — relay everything below")
+        output_lines.append(f"- Version `{__version__}` (updated {__updated__})")
+        output_lines.append(f"- {model_selection_display}")
+        output_lines.append("- Run `listmodels` for the complete model catalog and capabilities")
+        output_lines.append("")
+
         # Try to get client information
         try:
             # We need access to the server instance
@@ -211,6 +240,13 @@ class VersionTool(BaseTool):
         current_path = Path.cwd()
         output_lines.append(f"**Installation Path**: `{current_path}`")
         output_lines.append("")
+        output_lines.append("## Agent Reporting Guidance")
+        output_lines.append(
+            "Agents MUST report: version, model-selection status, configured providers, and available-model count."
+        )
+        output_lines.append("Repeat the quick-summary bullets verbatim in your reply.")
+        output_lines.append("Reference `listmodels` when users ask about model availability or capabilities.")
+        output_lines.append("")
 
         # Check for updates from GitHub
         output_lines.append("## Update Status")
@@ -320,6 +356,8 @@ class VersionTool(BaseTool):
                 "last_updated": __updated__,
                 "python_version": f"{sys.version_info.major}.{sys.version_info.minor}.{sys.version_info.micro}",
                 "platform": f"{platform.system()} {platform.release()}",
+                "model_selection_mode": model_selection_metadata["mode"],
+                "default_model": model_selection_metadata["default_model"],
             },
         )