aa

gurt/config.py

@@ -33,7 +33,7 @@ TAVILY_DEFAULT_MAX_RESULTS = int(os.getenv("TAVILY_DEFAULT_MAX_RESULTS", 5))
 TAVILY_DISABLE_ADVANCED = os.getenv("TAVILY_DISABLE_ADVANCED", "false").lower() == "true" # For cost control
 
 # --- Model Configuration ---
-DEFAULT_MODEL = os.getenv("GURT_DEFAULT_MODEL", "gemini-2.5-flash-preview-04-17")
+DEFAULT_MODEL = os.getenv("GURT_DEFAULT_MODEL", "projects/gurting/locations/us-central1/models/Gurt")
 FALLBACK_MODEL = os.getenv("GURT_FALLBACK_MODEL", "gemini-2.5-flash-preview-04-17")
 SAFETY_CHECK_MODEL = os.getenv("GURT_SAFETY_CHECK_MODEL", "gemini-2.5-flash-preview-04-17") # Use a Vertex AI model for safety checks
 
@@ -966,4 +966,4 @@ GURT_RESPONSES = [
     "Gurt!", "Gurt gurt!", "Gurt... gurt gurt.", "*gurts happily*",
     "*gurts sadly*", "*confused gurting*", "Gurt? Gurt gurt!", "GURT!",
     "gurt...", "Gurt gurt gurt!", "*aggressive gurting*"
-]
+]

gurt/extrtools.py

@@ -0,0 +1,69 @@
+#!/usr/bin/env python3
+"""
+extract_tools.py
+
+Usage:
+    python extract_tools.py path/to/your_module.py > tools.json
+
+Parses the given Python file for calls to:
+    tool_declarations.append(
+        generative_models.FunctionDeclaration(...)
+    )
+and outputs a JSON list of the kwargs passed to each FunctionDeclaration.
+"""
+
+import ast
+import json
+import sys
+
+def extract_function_declarations(source: str):
+    tree = ast.parse(source)
+    tools = []
+
+    for node in ast.walk(tree):
+        # look for expressions like: tool_declarations.append( generative_models.FunctionDeclaration(...) )
+        if (
+            isinstance(node, ast.Expr)
+            and isinstance(node.value, ast.Call)
+            and isinstance(node.value.func, ast.Attribute)
+            and node.value.func.attr == "append"
+            # ensure it's tool_declarations.append
+            and isinstance(node.value.func.value, ast.Name)
+            and node.value.func.value.id == "tool_declarations"
+            and node.value.args
+            and isinstance(node.value.args[0], ast.Call)
+        ):
+            decl_call = node.value.args[0]
+            # ensure it's generative_models.FunctionDeclaration(...)
+            if (
+                isinstance(decl_call.func, ast.Attribute)
+                and decl_call.func.attr == "FunctionDeclaration"
+            ):
+                tool_obj = {}
+                for kw in decl_call.keywords:
+                    # use ast.literal_eval to turn the AST node into a Python object
+                    try:
+                        value = ast.literal_eval(kw.value)
+                    except ValueError:
+                        # if something non-literal sneaks in, fallback to the raw source
+                        value = ast.get_source_segment(source, kw.value)
+                    tool_obj[kw.arg] = value
+                tools.append(tool_obj)
+
+    return tools
+
+def main():
+    if len(sys.argv) != 2:
+        print("Usage: python extract_tools.py path/to/your_module.py", file=sys.stderr)
+        sys.exit(1)
+
+    path = sys.argv[1]
+    with open(path, "r", encoding="utf-8") as f:
+        source = f.read()
+
+    tools = extract_function_declarations(source)
+    json.dump(tools, sys.stdout, indent=2)
+    sys.stdout.write("\n")
+
+if __name__ == "__main__":
+    main()