diff --git a/README.md b/README.md
index 1a87d1e..461bcf4 100644
--- a/README.md
+++ b/README.md
@@ -77,6 +77,10 @@ package schema
 
 #Classes: {
     MyCustomGAM: {
+        #meta: {
+            direction: "INOUT"
+            multithreaded: true
+        }
         Param1: int
         Param2?: string
         ...
diff --git a/specification.md b/specification.md
index d6f1b71..eac44ac 100644
--- a/specification.md
+++ b/specification.md
@@ -21,11 +21,12 @@ The executable should support the following subcommands:
 The LSP server should provide the following capabilities:
 
 - **Diagnostics**: Report syntax errors and validation issues.
+- **Incremental Sync**: Supports `textDocumentSync` kind 2 (Incremental) for better performance with large files.
 - **Hover Documentation**:
   - **Objects**: Display `CLASS::Name` and any associated docstrings.
   - **Signals**: Display `DataSource.Name TYPE (SIZE) [IN/OUT/INOUT]` along with docstrings.
   - **GAMs**: Show the list of States where the GAM is referenced.
-  - **Referenced Signals**: Show the list of GAMs where the signal is referenced.
+  - **Referenced Signals**: Show the list of GAMs where the signal is referenced (indicating Input/Output direction).
 - **Go to Definition**: Jump to the definition of a reference, supporting navigation across any file in the current project.
 - **Go to References**: Find usages of a node or field, supporting navigation across any file in the current project.
 - **Code Completion**: Autocomplete fields, values, and references.
@@ -173,6 +174,7 @@ The tool must build an index of the configuration to support LSP features and va
     - **Conditional Fields**: Validation of fields whose presence or value depends on the values of other fields within the same node or context.
   - **Schema Definition**:
     - Class validation rules must be defined in a separate schema file using the **CUE** language.
+    - **Metadata**: Class properties like direction (`#direction`) and multithreading support (`#multithreaded`) are stored within a `#meta` field in the class definition (e.g., `#meta: { direction: "IN", multithreaded: true }`).
     - **Project-Specific Classes**: Developers can define their own project-specific classes and corresponding validation rules, expanding the validation capabilities for their specific needs.
   - **Schema Loading**:
     - **Default Schema**: The tool should look for a default schema file `marte_schema.cue` in standard system locations:
@@ -218,6 +220,7 @@ The LSP and `check` command should report the following:
     - Field type mismatches.
     - Grammar errors (e.g., missing closing brackets).
     - **Invalid Function Reference**: Elements in the `Functions` array of a `State.Thread` must be valid references to defined GAM nodes.
+  - **Threading Violation**: A DataSource that is not marked as multithreaded (via `#meta.multithreaded`) is used by GAMs running in different threads within the same State.
 
 ## Logging
 
diff --git a/test/ast_test.go b/test/ast_test.go
new file mode 100644
index 0000000..39d2071
--- /dev/null
+++ b/test/ast_test.go
@@ -0,0 +1,94 @@
+package integration
+
+import (
+	"testing"
+
+	"github.com/marte-community/marte-dev-tools/internal/parser"
+)
+
+func TestASTCoverage(t *testing.T) {
+	pos := parser.Position{Line: 1, Column: 1}
+
+	var n parser.Node
+	// var d parser.Definition // Definition has unexported method, can't assign?
+	// Yes I can assign if I am using the interface type exported by parser.
+	// But I cannot call the method.
+	var d parser.Definition
+	var v parser.Value
+
+	// Field
+	f := &parser.Field{Position: pos}
+	n = f
+	d = f
+	if n.Pos() != pos {
+		t.Error("Field.Pos failed")
+	}
+	_ = d
+
+	// ObjectNode
+	o := &parser.ObjectNode{Position: pos}
+	n = o
+	d = o
+	if n.Pos() != pos {
+		t.Error("ObjectNode.Pos failed")
+	}
+
+	// StringValue
+	sv := &parser.StringValue{Position: pos}
+	n = sv
+	v = sv
+	if n.Pos() != pos {
+		t.Error("StringValue.Pos failed")
+	}
+	_ = v
+
+	// IntValue
+	iv := &parser.IntValue{Position: pos}
+	n = iv
+	v = iv
+	if n.Pos() != pos {
+		t.Error("IntValue.Pos failed")
+	}
+
+	// FloatValue
+	fv := &parser.FloatValue{Position: pos}
+	n = fv
+	v = fv
+	if n.Pos() != pos {
+		t.Error("FloatValue.Pos failed")
+	}
+
+	// BoolValue
+	bv := &parser.BoolValue{Position: pos}
+	n = bv
+	v = bv
+	if n.Pos() != pos {
+		t.Error("BoolValue.Pos failed")
+	}
+
+	// ReferenceValue
+	rv := &parser.ReferenceValue{Position: pos}
+	n = rv
+	v = rv
+	if n.Pos() != pos {
+		t.Error("ReferenceValue.Pos failed")
+	}
+
+	// ArrayValue
+	av := &parser.ArrayValue{Position: pos}
+	n = av
+	v = av
+	if n.Pos() != pos {
+		t.Error("ArrayValue.Pos failed")
+	}
+
+	// Package
+	pkg := &parser.Package{Position: pos}
+	// Package implements Node?
+	// ast.go: func (p *Package) Pos() Position { return p.Position }
+	// Yes.
+	n = pkg
+	if n.Pos() != pos {
+		t.Error("Package.Pos failed")
+	}
+}