Fermat MCP

Instructions

Implementation Reference

• fmcp/numpy_mcp/core/numerical_operation.py:5-167 (handler)

  The core handler function implementing the 'numpy_mcp_numerical_operation' tool logic using NumPy for various array and linear algebra operations.
  def numerical_operation( operation: str, a: Optional[Union[list, float, int]] = None, b: Optional[Union[list, float, int]] = None, shape: Optional[list] = None, new_shape: Optional[list] = None, axis: int = 0, q: Optional[float] = None, start: Optional[float] = None, stop: Optional[float] = None, step: Optional[float] = 1.0, num: Optional[int] = None, fill_value: Optional[float] = None, ) -> Union[list, float, int, dict]: """ Unified interface for numerical operations using numpy. Args: operation: The operation to perform. One of: - 'create_array': Convert list to array (a: list) - 'zeros': Array of zeros (shape: list) - 'ones': Array of ones (shape: list) - 'full': Array filled with value (shape: list, fill_value: float) - 'arange': Evenly spaced values (start: float, stop: float, step: float) - 'linspace': Evenly spaced samples (start: float, stop: float, num: int) - 'reshape': Reshape array (a: list, new_shape: list) - 'flatten': Flatten array (a: list) - 'concatenate': Join arrays (a: list[list], axis: int) - 'transpose': Transpose array (a: list) - 'stack': Stack arrays (a: list[list], axis: int) - 'add': Element-wise addition (a: list, b: list) - 'subtract': Element-wise subtraction (a: list, b: list) - 'multiply': Element-wise multiplication (a: list, b: list) - 'divide': Element-wise division (a: list, b: list) - 'power': Element-wise power (a: list, b: list) - 'abs_val': Absolute value (a: list) - 'exp': Exponential (a: list) - 'log': Natural logarithm (a: list) - 'sqrt': Square root (a: list) - 'sin': Sine (a: list) - 'cos': Cosine (a: list) - 'tan': Tangent (a: list) - 'mean': Mean (a: list) - 'median': Median (a: list) - 'std': Standard deviation (a: list) - 'var': Variance (a: list) - 'min_val': Minimum value (a: list) - 'max_val': Maximum value (a: list) - 'argmin': Index of minimum (a: list) - 'argmax': Index of maximum (a: list) - 'percentile': Percentile (a: list, q: float) - 'dot': Dot product (a: list, b: list) - 'matmul': Matrix multiplication (a: list, b: list) - 'inv': Matrix inverse (a: list) - 'det': Matrix determinant (a: list) - 'eig': Eigenvalues/vectors (a: list) - 'eigenvals': Eigenvalues only (a: list) - 'solve': Solve linear system (a: list, b: list) - 'svd': Singular Value Decomposition (a: list) a: First input array or value b: Second input array or value (for binary operations) shape: Shape of the output array (for creation operations) new_shape: New shape for reshape operation axis: Axis along which to perform operations (default: 0) q: Percentile value (0-100) for percentile operation start: Start value for arange/linspace stop: Stop value for arange/linspace step: Step size for arange (default: 1.0) num: Number of samples for linspace fill_value: Fill value for 'full' operation Returns: Result of the operation, type depends on operation """ a_array = np.array(a) if a is not None else None b_array = np.array(b) if b is not None else None if operation == "create_array": return a_array.tolist() elif operation == "zeros": return np.zeros(shape).tolist() elif operation == "ones": return np.ones(shape).tolist() elif operation == "full": return np.full(shape, fill_value).tolist() elif operation == "arange": return np.arange(start, stop, step).tolist() elif operation == "linspace": return np.linspace(start, stop, num).tolist() elif operation == "reshape": return a_array.reshape(new_shape).tolist() elif operation == "flatten": return a_array.ravel().tolist() elif operation == "concatenate": return np.concatenate([np.array(arr) for arr in a], axis=axis).tolist() elif operation == "transpose": return a_array.T.tolist() elif operation == "stack": return np.stack([np.array(arr) for arr in a], axis=axis).tolist() elif operation == "add": return (a_array + b_array).tolist() elif operation == "subtract": return (a_array - b_array).tolist() elif operation == "multiply": return (a_array * b_array).tolist() elif operation == "divide": return (a_array / b_array).tolist() elif operation == "power": return np.power(a_array, b_array).tolist() elif operation == "abs_val": return np.abs(a_array).tolist() elif operation == "exp": return np.exp(a_array).tolist() elif operation == "log": return np.log(a_array).tolist() elif operation == "sqrt": return np.sqrt(a_array).tolist() elif operation == "sin": return np.sin(a_array).tolist() elif operation == "cos": return np.cos(a_array).tolist() elif operation == "tan": return np.tan(a_array).tolist() elif operation == "mean": return float(np.mean(a_array)) elif operation == "median": return float(np.median(a_array)) elif operation == "std": return float(np.std(a_array)) elif operation == "var": return float(np.var(a_array)) elif operation == "min_val": return float(np.min(a_array)) elif operation == "max_val": return float(np.max(a_array)) elif operation == "argmin": return int(np.argmin(a_array)) elif operation == "argmax": return int(np.argmax(a_array)) elif operation == "percentile": return float(np.percentile(a_array, q)) elif operation == "dot": return float(np.dot(a_array, b_array)) elif operation == "matmul": return np.matmul(a_array, b_array).tolist() elif operation == "inv": return np.linalg.inv(a_array).tolist() elif operation == "det": return float(np.linalg.det(a_array)) elif operation == "eig": vals, vecs = np.linalg.eig(a_array) return {"eigenvalues": vals.tolist(), "eigenvectors": vecs.tolist()} elif operation == "eigenvals": vals = np.linalg.eigvals(a_array) return vals.tolist() elif operation == "solve": return np.linalg.solve(a_array, b_array).tolist() elif operation == "svd": U, S, Vt = np.linalg.svd(a_array) return {"U": U.tolist(), "S": S.tolist(), "Vt": Vt.tolist()} else: raise ValueError(f"Unknown operation: {operation}")
• fmcp/numpy_mcp/server.py:14-17 (registration)

  Registers the numerical_operation function as the 'numpy_mcp_numerical_operation' tool in the FastMCP server.
  numpy_mcp.tool( numerical_operation, description="Do numerical operation like add, sub, mul, div, power, abs, exp, log, sqrt, sin, cos, tan, mean, median, std, var, min, max, argmin, argmax, percentile, dot, matmul, inv, det, eig, solve, svd", )
• fmcp/numpy_mcp/core/numerical_operation.py:5-78 (schema)

  Type hints and comprehensive docstring defining the input schema and supported operations for the tool.
  def numerical_operation( operation: str, a: Optional[Union[list, float, int]] = None, b: Optional[Union[list, float, int]] = None, shape: Optional[list] = None, new_shape: Optional[list] = None, axis: int = 0, q: Optional[float] = None, start: Optional[float] = None, stop: Optional[float] = None, step: Optional[float] = 1.0, num: Optional[int] = None, fill_value: Optional[float] = None, ) -> Union[list, float, int, dict]: """ Unified interface for numerical operations using numpy. Args: operation: The operation to perform. One of: - 'create_array': Convert list to array (a: list) - 'zeros': Array of zeros (shape: list) - 'ones': Array of ones (shape: list) - 'full': Array filled with value (shape: list, fill_value: float) - 'arange': Evenly spaced values (start: float, stop: float, step: float) - 'linspace': Evenly spaced samples (start: float, stop: float, num: int) - 'reshape': Reshape array (a: list, new_shape: list) - 'flatten': Flatten array (a: list) - 'concatenate': Join arrays (a: list[list], axis: int) - 'transpose': Transpose array (a: list) - 'stack': Stack arrays (a: list[list], axis: int) - 'add': Element-wise addition (a: list, b: list) - 'subtract': Element-wise subtraction (a: list, b: list) - 'multiply': Element-wise multiplication (a: list, b: list) - 'divide': Element-wise division (a: list, b: list) - 'power': Element-wise power (a: list, b: list) - 'abs_val': Absolute value (a: list) - 'exp': Exponential (a: list) - 'log': Natural logarithm (a: list) - 'sqrt': Square root (a: list) - 'sin': Sine (a: list) - 'cos': Cosine (a: list) - 'tan': Tangent (a: list) - 'mean': Mean (a: list) - 'median': Median (a: list) - 'std': Standard deviation (a: list) - 'var': Variance (a: list) - 'min_val': Minimum value (a: list) - 'max_val': Maximum value (a: list) - 'argmin': Index of minimum (a: list) - 'argmax': Index of maximum (a: list) - 'percentile': Percentile (a: list, q: float) - 'dot': Dot product (a: list, b: list) - 'matmul': Matrix multiplication (a: list, b: list) - 'inv': Matrix inverse (a: list) - 'det': Matrix determinant (a: list) - 'eig': Eigenvalues/vectors (a: list) - 'eigenvals': Eigenvalues only (a: list) - 'solve': Solve linear system (a: list, b: list) - 'svd': Singular Value Decomposition (a: list) a: First input array or value b: Second input array or value (for binary operations) shape: Shape of the output array (for creation operations) new_shape: New shape for reshape operation axis: Axis along which to perform operations (default: 0) q: Percentile value (0-100) for percentile operation start: Start value for arange/linspace stop: Stop value for arange/linspace step: Step size for arange (default: 1.0) num: Number of samples for linspace fill_value: Fill value for 'full' operation Returns: Result of the operation, type depends on operation """