feat(resolvemcp): add support for join-column sorting in cursor pagination

* Enhance getCursorFilter to accept join clauses for sorting
* Update resolveColumn to handle joined columns
* Modify tests to validate new join functionality

pkg/resolvemcp/cursor.go

@@ -18,11 +18,13 @@ const (
 )
 
 // getCursorFilter generates a SQL EXISTS subquery for cursor-based pagination.
+// expandJoins is an optional map[alias]string of JOIN clauses for join-column sort support.
 func getCursorFilter(
 	tableName string,
 	pkName string,
 	modelColumns []string,
 	options common.RequestOptions,
+	expandJoins map[string]string,
 ) (string, error) {
 	fullTableName := tableName
 	if strings.Contains(tableName, ".") {
@@ -40,6 +42,7 @@ func getCursorFilter(
 	}
 
 	var whereClauses []string
+	joinSQL := ""
 	reverse := direction < 0
 
 	for _, s := range sortItems {
@@ -57,12 +60,27 @@ func getCursorFilter(
 			desc = !desc
 		}
 
-		cursorCol, targetCol, err := resolveCursorColumn(field, prefix, tableName, modelColumns)
+		cursorCol, targetCol, isJoin, err := resolveCursorColumn(field, prefix, tableName, modelColumns)
 		if err != nil {
 			logger.Warn("Skipping invalid sort column %q: %v", col, err)
 			continue
 		}
 
+		if isJoin {
+			if expandJoins != nil {
+				if joinClause, ok := expandJoins[prefix]; ok {
+					jSQL, cRef := rewriteCursorJoin(joinClause, tableName, prefix)
+					joinSQL = jSQL
+					cursorCol = cRef + "." + field
+					targetCol = prefix + "." + field
+				}
+			}
+			if cursorCol == "" {
+				logger.Warn("Skipping cursor sort column %q: join alias %q not in expandJoins", col, prefix)
+				continue
+			}
+		}
+
 		op := "<"
 		if desc {
 			op = ">"
@@ -79,10 +97,12 @@ func getCursorFilter(
 	query := fmt.Sprintf(`EXISTS (
   SELECT 1
   FROM %s cursor_select
+  %s
   WHERE cursor_select.%s = %s
     AND (%s)
 )`,
 		fullTableName,
+		joinSQL,
 		pkName,
 		cursorID,
 		orSQL,
@@ -101,26 +121,34 @@ func getActiveCursor(options common.RequestOptions) (id string, direction cursor
 	return "", 0
 }
 
-func resolveCursorColumn(field, prefix, tableName string, modelColumns []string) (cursorCol, targetCol string, err error) {
+func resolveCursorColumn(field, prefix, tableName string, modelColumns []string) (cursorCol, targetCol string, isJoin bool, err error) {
 	if strings.Contains(field, "->") {
-		return "cursor_select." + field, tableName + "." + field, nil
+		return "cursor_select." + field, tableName + "." + field, false, nil
 	}
 
 	if modelColumns != nil {
 		for _, col := range modelColumns {
 			if strings.EqualFold(col, field) {
-				return "cursor_select." + field, tableName + "." + field, nil
+				return "cursor_select." + field, tableName + "." + field, false, nil
 			}
 		}
 	} else {
-		return "cursor_select." + field, tableName + "." + field, nil
+		return "cursor_select." + field, tableName + "." + field, false, nil
 	}
 
 	if prefix != "" && prefix != tableName {
-		return "", "", fmt.Errorf("joined columns not supported in cursor pagination: %s", field)
+		return "", "", true, nil
 	}
 
-	return "", "", fmt.Errorf("invalid column: %s", field)
+	return "", "", false, fmt.Errorf("invalid column: %s", field)
+}
+
+func rewriteCursorJoin(joinClause, mainTable, alias string) (joinSQL, cursorAlias string) {
+	joinSQL = strings.ReplaceAll(joinClause, mainTable+".", "cursor_select.")
+	cursorAlias = "cursor_select_" + alias
+	joinSQL = strings.ReplaceAll(joinSQL, " "+alias+" ", " "+cursorAlias+" ")
+	joinSQL = strings.ReplaceAll(joinSQL, " "+alias+".", " "+cursorAlias+".")
+	return joinSQL, cursorAlias
 }
 
 func buildCursorPriorityChain(clauses []string) string {

pkg/resolvemcp/handler.go

@@ -191,7 +191,8 @@ func (h *Handler) executeRead(ctx context.Context, schema, entity, id string, op
 			options.Sort = []common.SortOption{{Column: pkName, Direction: "ASC"}}
 		}
 
-		cursorFilter, err := getCursorFilter(tableName, pkName, modelColumns, options)
+		// expandJoins is empty for resolvemcp — no custom SQL join support yet
+		cursorFilter, err := getCursorFilter(tableName, pkName, modelColumns, options, nil)
 		if err != nil {
 			return nil, nil, fmt.Errorf("cursor error: %w", err)
 		}

pkg/resolvemcp/tools.go

@@ -4,12 +4,14 @@ import (
 	"context"
 	"encoding/json"
 	"fmt"
+	"reflect"
 	"strings"
 
 	"github.com/mark3labs/mcp-go/mcp"
 
 	"github.com/bitechdev/ResolveSpec/pkg/common"
 	"github.com/bitechdev/ResolveSpec/pkg/logger"
+	"github.com/bitechdev/ResolveSpec/pkg/reflection"
 )
 
 // toolName builds the MCP tool name for a given operation and model.
@@ -22,54 +24,268 @@ func toolName(operation, schema, entity string) string {
 
 // registerModelTools registers the four CRUD tools and resource for a model.
 func registerModelTools(h *Handler, schema, entity string, model interface{}) {
-	registerReadTool(h, schema, entity)
-	registerCreateTool(h, schema, entity)
-	registerUpdateTool(h, schema, entity)
-	registerDeleteTool(h, schema, entity)
-	registerModelResource(h, schema, entity)
+	info := buildModelInfo(schema, entity, model)
+	registerReadTool(h, schema, entity, info)
+	registerCreateTool(h, schema, entity, info)
+	registerUpdateTool(h, schema, entity, info)
+	registerDeleteTool(h, schema, entity, info)
+	registerModelResource(h, schema, entity, info)
 
-	logger.Info("[resolvemcp] Registered MCP tools for %s.%s", schema, entity)
+	logger.Info("[resolvemcp] Registered MCP tools for %s", info.fullName)
+}
+
+// --------------------------------------------------------------------------
+// Model introspection
+// --------------------------------------------------------------------------
+
+// modelInfo holds pre-computed metadata for a model used in tool descriptions.
+type modelInfo struct {
+	fullName      string // e.g. "public.users"
+	pkName        string // e.g. "id"
+	columns       []columnInfo
+	relationNames []string
+	schemaDoc     string // formatted multi-line schema listing
+}
+
+type columnInfo struct {
+	jsonName  string
+	sqlName   string
+	goType    string
+	sqlType   string
+	isPrimary bool
+	isUnique  bool
+	isFK      bool
+	nullable  bool
+}
+
+// buildModelInfo extracts column metadata and pre-builds the schema documentation string.
+func buildModelInfo(schema, entity string, model interface{}) modelInfo {
+	info := modelInfo{
+		fullName: buildModelName(schema, entity),
+		pkName:   reflection.GetPrimaryKeyName(model),
+	}
+
+	// Unwrap to base struct type
+	modelType := reflect.TypeOf(model)
+	for modelType != nil && (modelType.Kind() == reflect.Ptr || modelType.Kind() == reflect.Slice) {
+		modelType = modelType.Elem()
+	}
+	if modelType == nil || modelType.Kind() != reflect.Struct {
+		return info
+	}
+
+	details := reflection.GetModelColumnDetail(reflect.New(modelType).Elem())
+
+	for _, d := range details {
+		// Derive the JSON name from the struct field
+		jsonName := fieldJSONName(modelType, d.Name)
+		if jsonName == "" || jsonName == "-" {
+			continue
+		}
+
+		// Skip relation fields (slice or user-defined struct that isn't time.Time).
+		fieldType, found := modelType.FieldByName(d.Name)
+		if found {
+			ft := fieldType.Type
+			if ft.Kind() == reflect.Ptr {
+				ft = ft.Elem()
+			}
+			isUserStruct := ft.Kind() == reflect.Struct && ft.Name() != "Time" && ft.PkgPath() != ""
+			if ft.Kind() == reflect.Slice || isUserStruct {
+				info.relationNames = append(info.relationNames, jsonName)
+				continue
+			}
+		}
+
+		sqlName := d.SQLName
+		if sqlName == "" {
+			sqlName = jsonName
+		}
+
+		// Derive Go type name, unwrapping pointer if needed.
+		goType := d.DataType
+		if goType == "" && found {
+			ft := fieldType.Type
+			for ft.Kind() == reflect.Ptr {
+				ft = ft.Elem()
+			}
+			goType = ft.Name()
+		}
+
+		// isPrimary: use both the GORM-tag detection and a name comparison against
+		// the known primary key (handles camelCase "primaryKey" tags correctly).
+		isPrimary := d.SQLKey == "primary_key" ||
+			(info.pkName != "" && (sqlName == info.pkName || jsonName == info.pkName))
+
+		ci := columnInfo{
+			jsonName:  jsonName,
+			sqlName:   sqlName,
+			goType:    goType,
+			sqlType:   d.SQLDataType,
+			isPrimary: isPrimary,
+			isUnique:  d.SQLKey == "unique" || d.SQLKey == "uniqueindex",
+			isFK:      d.SQLKey == "foreign_key",
+			nullable:  d.Nullable,
+		}
+		info.columns = append(info.columns, ci)
+	}
+
+	info.schemaDoc = buildSchemaDoc(info)
+	return info
+}
+
+// fieldJSONName returns the JSON tag name for a struct field, falling back to the field name.
+func fieldJSONName(modelType reflect.Type, fieldName string) string {
+	field, ok := modelType.FieldByName(fieldName)
+	if !ok {
+		return fieldName
+	}
+	tag := field.Tag.Get("json")
+	if tag == "" {
+		return fieldName
+	}
+	parts := strings.SplitN(tag, ",", 2)
+	if parts[0] == "" {
+		return fieldName
+	}
+	return parts[0]
+}
+
+// buildSchemaDoc builds a human-readable column listing for inclusion in tool descriptions.
+func buildSchemaDoc(info modelInfo) string {
+	if len(info.columns) == 0 {
+		return ""
+	}
+
+	var sb strings.Builder
+	sb.WriteString("Columns:\n")
+	for _, c := range info.columns {
+		line := fmt.Sprintf("  • %s", c.jsonName)
+
+		typeDesc := c.goType
+		if c.sqlType != "" {
+			typeDesc = c.sqlType
+		}
+		if typeDesc != "" {
+			line += fmt.Sprintf(" (%s)", typeDesc)
+		}
+
+		var flags []string
+		if c.isPrimary {
+			flags = append(flags, "primary key")
+		}
+		if c.isUnique {
+			flags = append(flags, "unique")
+		}
+		if c.isFK {
+			flags = append(flags, "foreign key")
+		}
+		if !c.nullable && !c.isPrimary {
+			flags = append(flags, "not null")
+		} else if c.nullable {
+			flags = append(flags, "nullable")
+		}
+		if len(flags) > 0 {
+			line += " — " + strings.Join(flags, ", ")
+		}
+
+		sb.WriteString(line + "\n")
+	}
+
+	if len(info.relationNames) > 0 {
+		sb.WriteString("Relations (preloadable): " + strings.Join(info.relationNames, ", ") + "\n")
+	}
+
+	return sb.String()
+}
+
+// columnNameList returns a comma-separated list of JSON column names (for descriptions).
+func columnNameList(cols []columnInfo) string {
+	names := make([]string, len(cols))
+	for i, c := range cols {
+		names[i] = c.jsonName
+	}
+	return strings.Join(names, ", ")
+}
+
+// writableColumnNames returns JSON names for all non-primary-key columns.
+func writableColumnNames(cols []columnInfo) []string {
+	var names []string
+	for _, c := range cols {
+		if !c.isPrimary {
+			names = append(names, c.jsonName)
+		}
+	}
+	return names
 }
 
 // --------------------------------------------------------------------------
 // Read tool
 // --------------------------------------------------------------------------
 
-func registerReadTool(h *Handler, schema, entity string) {
+func registerReadTool(h *Handler, schema, entity string, info modelInfo) {
 	name := toolName("read", schema, entity)
-	description := fmt.Sprintf("Read records from %s", buildModelName(schema, entity))
+
+	var descParts []string
+	descParts = append(descParts, fmt.Sprintf("Read records from the '%s' database table.", info.fullName))
+	if info.pkName != "" {
+		descParts = append(descParts, fmt.Sprintf("Primary key: '%s'. Pass it via 'id' to fetch a single record.", info.pkName))
+	}
+	if info.schemaDoc != "" {
+		descParts = append(descParts, info.schemaDoc)
+	}
+	descParts = append(descParts,
+		"Pagination: use 'limit'/'offset' for offset-based paging, or 'cursor_forward'/'cursor_backward' (pass the primary key value of the last/first record on the current page) for cursor-based paging.",
+		"Filtering: each filter object requires 'column' (JSON field name) and 'operator'. Supported operators: = != > < >= <= like ilike in is_null is_not_null. Combine with 'logic_operator': AND (default) or OR.",
+		"Sorting: each sort object requires 'column' and 'direction' (asc or desc).",
+	)
+	if len(info.relationNames) > 0 {
+		descParts = append(descParts, fmt.Sprintf("Preloadable relations: %s. Pass relation name in 'preloads'.", strings.Join(info.relationNames, ", ")))
+	}
+
+	description := strings.Join(descParts, "\n\n")
+
+	filterDesc := fmt.Sprintf(`Array of filter objects. Example: [{"column":"status","operator":"=","value":"active"},{"column":"age","operator":">","value":18,"logic_operator":"AND"}]`)
+	if len(info.columns) > 0 {
+		filterDesc += fmt.Sprintf(" Available columns: %s.", columnNameList(info.columns))
+	}
+
+	sortDesc := `Array of sort objects. Example: [{"column":"created_at","direction":"desc"}]`
+	if len(info.columns) > 0 {
+		sortDesc += fmt.Sprintf(" Available columns: %s.", columnNameList(info.columns))
+	}
 
 	tool := mcp.NewTool(name,
 		mcp.WithDescription(description),
 		mcp.WithString("id",
-			mcp.Description("Primary key of a single record to fetch (optional)"),
+			mcp.Description(fmt.Sprintf("Primary key (%s) of a single record to fetch. Omit to return multiple records.", info.pkName)),
 		),
 		mcp.WithNumber("limit",
-			mcp.Description("Maximum number of records to return"),
+			mcp.Description("Maximum number of records to return per page. Recommended: 10–100."),
 		),
 		mcp.WithNumber("offset",
-			mcp.Description("Number of records to skip"),
+			mcp.Description("Number of records to skip (for offset-based pagination). Use with 'limit'."),
 		),
 		mcp.WithString("cursor_forward",
-			mcp.Description("Cursor value for the next page (primary key of last record on current page)"),
+			mcp.Description(fmt.Sprintf("Cursor for the next page: pass the '%s' value of the last record on the current page. Requires 'sort' to be set.", info.pkName)),
 		),
 		mcp.WithString("cursor_backward",
-			mcp.Description("Cursor value for the previous page"),
+			mcp.Description(fmt.Sprintf("Cursor for the previous page: pass the '%s' value of the first record on the current page. Requires 'sort' to be set.", info.pkName)),
 		),
 		mcp.WithArray("columns",
-			mcp.Description("List of column names to include in the result"),
+			mcp.Description(fmt.Sprintf("Columns to include in the result. Omit to return all columns. Available: %s.", columnNameList(info.columns))),
 		),
 		mcp.WithArray("omit_columns",
-			mcp.Description("List of column names to exclude from the result"),
+			mcp.Description(fmt.Sprintf("Columns to exclude from the result. Available: %s.", columnNameList(info.columns))),
 		),
 		mcp.WithArray("filters",
-			mcp.Description(`Array of filter objects. Each object: {"column":"name","operator":"=","value":"val","logic_operator":"AND|OR"}. Operators: =, !=, >, <, >=, <=, like, ilike, in, is_null, is_not_null`),
+			mcp.Description(filterDesc),
 		),
 		mcp.WithArray("sort",
-			mcp.Description(`Array of sort objects. Each object: {"column":"name","direction":"asc|desc"}`),
+			mcp.Description(sortDesc),
 		),
 		mcp.WithArray("preloads",
-			mcp.Description(`Array of relation preload objects. Each object: {"relation":"RelationName","columns":["col1"]}`),
+			mcp.Description(buildPreloadDesc(info)),
 		),
 	)
 
@@ -91,18 +307,52 @@ func registerReadTool(h *Handler, schema, entity string) {
 	})
 }
 
+func buildPreloadDesc(info modelInfo) string {
+	if len(info.relationNames) == 0 {
+		return `Array of relation preload objects. Each object: {"relation":"RelationName"}. No relations defined on this model.`
+	}
+	return fmt.Sprintf(
+		`Array of relation preload objects. Each object: {"relation":"RelationName","columns":["col1","col2"]}. Available relations: %s.`,
+		strings.Join(info.relationNames, ", "),
+	)
+}
+
 // --------------------------------------------------------------------------
 // Create tool
 // --------------------------------------------------------------------------
 
-func registerCreateTool(h *Handler, schema, entity string) {
+func registerCreateTool(h *Handler, schema, entity string, info modelInfo) {
 	name := toolName("create", schema, entity)
-	description := fmt.Sprintf("Create one or more records in %s", buildModelName(schema, entity))
+
+	writable := writableColumnNames(info.columns)
+
+	var descParts []string
+	descParts = append(descParts, fmt.Sprintf("Create one or more new records in the '%s' table.", info.fullName))
+	if len(writable) > 0 {
+		descParts = append(descParts, fmt.Sprintf("Writable fields: %s.", strings.Join(writable, ", ")))
+	}
+	if info.pkName != "" {
+		descParts = append(descParts, fmt.Sprintf("The primary key ('%s') is typically auto-generated — omit it unless you need to supply it explicitly.", info.pkName))
+	}
+	descParts = append(descParts,
+		"Pass a single JSON object to 'data' to create one record. Pass an array of objects to create multiple records in a single transaction (all succeed or all fail).",
+	)
+	if info.schemaDoc != "" {
+		descParts = append(descParts, info.schemaDoc)
+	}
+
+	description := strings.Join(descParts, "\n\n")
+
+	dataDesc := "Record fields to create."
+	if len(writable) > 0 {
+		dataDesc += fmt.Sprintf(" Writable fields: %s.", strings.Join(writable, ", "))
+	}
+	dataDesc += " Pass a single object or an array of objects."
 
 	tool := mcp.NewTool(name,
 		mcp.WithDescription(description),
 		mcp.WithObject("data",
-			mcp.Description("Record fields to create (single object), or pass an array as the 'items' key"),
+			mcp.Description(dataDesc),
 			mcp.Required(),
 		),
 	)
@@ -130,17 +380,42 @@ func registerCreateTool(h *Handler, schema, entity string) {
 // Update tool
 // --------------------------------------------------------------------------
 
-func registerUpdateTool(h *Handler, schema, entity string) {
+func registerUpdateTool(h *Handler, schema, entity string, info modelInfo) {
 	name := toolName("update", schema, entity)
-	description := fmt.Sprintf("Update an existing record in %s", buildModelName(schema, entity))
+
+	writable := writableColumnNames(info.columns)
+
+	var descParts []string
+	descParts = append(descParts, fmt.Sprintf("Update an existing record in the '%s' table.", info.fullName))
+	if info.pkName != "" {
+		descParts = append(descParts, fmt.Sprintf("Identify the record by its primary key ('%s') via the 'id' argument or by including '%s' inside 'data'.", info.pkName, info.pkName))
+	}
+	if len(writable) > 0 {
+		descParts = append(descParts, fmt.Sprintf("Updatable fields: %s.", strings.Join(writable, ", ")))
+	}
+	descParts = append(descParts,
+		"Only non-null, non-empty fields in 'data' are applied — existing values are preserved for fields you omit. Returns the merged record as stored.",
+	)
+	if info.schemaDoc != "" {
+		descParts = append(descParts, info.schemaDoc)
+	}
+
+	description := strings.Join(descParts, "\n\n")
+
+	idDesc := fmt.Sprintf("Primary key ('%s') of the record to update. Can also be included inside 'data'.", info.pkName)
+
+	dataDesc := "Fields to update (non-null, non-empty values are merged into the existing record)."
+	if len(writable) > 0 {
+		dataDesc += fmt.Sprintf(" Updatable fields: %s.", strings.Join(writable, ", "))
+	}
 
 	tool := mcp.NewTool(name,
 		mcp.WithDescription(description),
 		mcp.WithString("id",
-			mcp.Description("Primary key of the record to update"),
+			mcp.Description(idDesc),
 		),
 		mcp.WithObject("data",
-			mcp.Description("Fields to update (non-null fields will be merged into the existing record)"),
+			mcp.Description(dataDesc),
 			mcp.Required(),
 		),
 	)
@@ -174,14 +449,23 @@ func registerUpdateTool(h *Handler, schema, entity string) {
 // Delete tool
 // --------------------------------------------------------------------------
 
-func registerDeleteTool(h *Handler, schema, entity string) {
+func registerDeleteTool(h *Handler, schema, entity string, info modelInfo) {
 	name := toolName("delete", schema, entity)
-	description := fmt.Sprintf("Delete a record from %s by primary key", buildModelName(schema, entity))
+
+	descParts := []string{
+		fmt.Sprintf("Delete a record from the '%s' table by its primary key.", info.fullName),
+	}
+	if info.pkName != "" {
+		descParts = append(descParts, fmt.Sprintf("Pass the '%s' value of the record to delete via the 'id' argument.", info.pkName))
+	}
+	descParts = append(descParts, "Returns the deleted record. This operation is irreversible.")
+
+	description := strings.Join(descParts, " ")
 
 	tool := mcp.NewTool(name,
 		mcp.WithDescription(description),
 		mcp.WithString("id",
-			mcp.Description("Primary key of the record to delete"),
+			mcp.Description(fmt.Sprintf("Primary key ('%s') of the record to delete.", info.pkName)),
 			mcp.Required(),
 		),
 	)
@@ -206,17 +490,23 @@ func registerDeleteTool(h *Handler, schema, entity string) {
 // Resource registration
 // --------------------------------------------------------------------------
 
-func registerModelResource(h *Handler, schema, entity string) {
-	resourceURI := buildModelName(schema, entity)
-	displayName := entity
-	if schema != "" {
-		displayName = schema + "." + entity
+func registerModelResource(h *Handler, schema, entity string, info modelInfo) {
+	resourceURI := info.fullName
+
+	var resourceDesc strings.Builder
+	resourceDesc.WriteString(fmt.Sprintf("Database table: %s", info.fullName))
+	if info.pkName != "" {
+		resourceDesc.WriteString(fmt.Sprintf(" (primary key: %s)", info.pkName))
+	}
+	if info.schemaDoc != "" {
+		resourceDesc.WriteString("\n\n")
+		resourceDesc.WriteString(info.schemaDoc)
 	}
 
 	resource := mcp.NewResource(
 		resourceURI,
-		displayName,
-		mcp.WithResourceDescription(fmt.Sprintf("Database table: %s", displayName)),
+		entity,
+		mcp.WithResourceDescription(resourceDesc.String()),
 		mcp.WithMIMEType("application/json"),
 	)
 
@@ -256,7 +546,6 @@ func registerModelResource(h *Handler, schema, entity string) {
 func parseRequestOptions(args map[string]interface{}) common.RequestOptions {
 	options := common.RequestOptions{}
 
-	// limit
 	if v, ok := args["limit"]; ok {
 		switch n := v.(type) {
 		case float64:
@@ -267,7 +556,6 @@ func parseRequestOptions(args map[string]interface{}) common.RequestOptions {
 		}
 	}
 
-	// offset
 	if v, ok := args["offset"]; ok {
 		switch n := v.(type) {
 		case float64:
@@ -278,7 +566,6 @@ func parseRequestOptions(args map[string]interface{}) common.RequestOptions {
 		}
 	}
 
-	// cursor_forward / cursor_backward
 	if v, ok := args["cursor_forward"].(string); ok {
 		options.CursorForward = v
 	}
@@ -286,19 +573,10 @@ func parseRequestOptions(args map[string]interface{}) common.RequestOptions {
 		options.CursorBackward = v
 	}
 
-	// columns
 	options.Columns = parseStringArray(args["columns"])
-
-	// omit_columns
 	options.OmitColumns = parseStringArray(args["omit_columns"])
-
-	// filters — marshal each item and unmarshal into FilterOption
 	options.Filters = parseFilters(args["filters"])
-
-	// sort
 	options.Sort = parseSortOptions(args["sort"])
-
-	// preloads
 	options.Preload = parsePreloadOptions(args["preloads"])
 
 	return options
@@ -342,7 +620,6 @@ func parseFilters(raw interface{}) []common.FilterOption {
 		if f.Column == "" || f.Operator == "" {
 			continue
 		}
-		// Normalise logic operator
 		if strings.EqualFold(f.LogicOperator, "or") {
 			f.LogicOperator = "OR"
 		} else {