fix(server): skip oversized files in AST walk, fix line count

Two AST-summary/query path fixes plus colocated QF1012 style fixups:
- runASTQueryWalk now checks info.Size() > MaxFileSize before os.ReadFile, avoiding a full read of files it will discard.
- generateASTSummary used len(splitLines(...)) for the '(N lines)' header, over-counting newline-terminated files by one (phantom trailing element). New countLines() helper counts via bytes.Count. Adds TestCountLines.

internal/server/handlers_ast.go

@@ -137,11 +137,13 @@ func (s *Server) runASTQueryWalk(ctx context.Context, p *astQueryParams, exts []
 				return nil
 			}
 
-			content, err := os.ReadFile(path)
-			if err != nil {
+			// Skip oversized files before reading them off disk.
+			if info.Size() > s.cfg.MaxFileSize {
 				return nil
 			}
-			if int64(len(content)) > s.cfg.MaxFileSize {
+
+			content, err := os.ReadFile(path)
+			if err != nil {
 				return nil
 			}
 
@@ -260,12 +262,12 @@ func (s *Server) generateASTSummary(ctx context.Context, path string, content []
 		}
 	}
 
-	sb.WriteString(fmt.Sprintf("**%s** (%d lines, %s)\n\n", relPath, len(splitLines(string(content))), lang))
+	fmt.Fprintf(&sb, "**%s** (%d lines, %s)\n\n", relPath, countLines(content), lang)
 	sb.WriteString("Symbols:\n")
 
 	for _, sym := range symbols {
 		kindStr := symbolKindIcon(sym.Kind)
-		sb.WriteString(fmt.Sprintf("  %s %s  L%d\n", kindStr, sym.Name, sym.Location.Line))
+		fmt.Fprintf(&sb, "  %s %s  L%d\n", kindStr, sym.Name, sym.Location.Line)
 	}
 
 	return sb.String()

internal/server/handlers_file.go

@@ -3,6 +3,7 @@ package server
 
 import (
 	"bufio"
+	"bytes"
 	"context"
 	"fmt"
 	"os"
@@ -184,7 +185,7 @@ func (s *Server) handleFileRead(ctx context.Context, request mcp.CallToolRequest
 			}
 			result, err := s.readOneFile(ctx, request, p)
 			if err != nil {
-				output.WriteString(fmt.Sprintf("--- %s ---\n[error: %s]\n", p, errors.SanitizeError(err)))
+				fmt.Fprintf(&output, "--- %s ---\n[error: %s]\n", p, errors.SanitizeError(err))
 				continue
 			}
 			// Extract etag from result footer for dedup check.
@@ -192,12 +193,12 @@ func (s *Server) handleFileRead(ctx context.Context, request mcp.CallToolRequest
 			if etag != "" {
 				if firstPath, seen := seenEtag[etag]; seen {
 					// Duplicate content: emit pointer instead of full content.
-					output.WriteString(fmt.Sprintf("--- %s ---\n[duplicate of %s, etag: %s]\n", p, firstPath, etag))
+					fmt.Fprintf(&output, "--- %s ---\n[duplicate of %s, etag: %s]\n", p, firstPath, etag)
 					continue
 				}
 				seenEtag[etag] = p
 			}
-			output.WriteString(fmt.Sprintf("--- %s ---\n%s", p, result))
+			fmt.Fprintf(&output, "--- %s ---\n%s", p, result)
 		}
 		return mcp.NewToolResultText(output.String()), nil
 	}
@@ -571,6 +572,21 @@ func extractEtag(result string) string {
 // splitLines splits a string into lines.
 // For large files (> 1MB), uses bufio.Scanner which is more memory efficient.
 // For smaller files, uses simple string split which is faster.
+// countLines returns the number of source lines in content. Unlike
+// len(splitLines(...)), it does not over-count newline-terminated files
+// (a trailing "\n" does not introduce a phantom final line) and avoids
+// materializing the line slice.
+func countLines(content []byte) int {
+	if len(content) == 0 {
+		return 0
+	}
+	n := bytes.Count(content, []byte{'\n'})
+	if content[len(content)-1] != '\n' {
+		n++
+	}
+	return n
+}
+
 func splitLines(s string) []string {
 	const largeSizeThreshold = 1024 * 1024 // 1MB
 

internal/server/server_test.go

@@ -415,6 +415,32 @@ func TestSplitLinesLargeFileNoTrailingNewline(t *testing.T) {
 	}
 }
 
+// TestCountLines verifies countLines reports source line counts without the
+// phantom trailing line that len(splitLines(...)) produces for newline-terminated
+// files.
+func TestCountLines(t *testing.T) {
+	tests := []struct {
+		name    string
+		content string
+		want    int
+	}{
+		{name: "empty", content: "", want: 0},
+		{name: "single line no newline", content: "a", want: 1},
+		{name: "single line trailing newline", content: "a\n", want: 1},
+		{name: "two lines trailing newline", content: "a\nb\n", want: 2},
+		{name: "two lines no trailing newline", content: "a\nb", want: 2},
+		{name: "blank lines", content: "\n\n\n", want: 3},
+		{name: "crlf trailing", content: "a\r\nb\r\n", want: 2},
+	}
+	for _, tt := range tests {
+		t.Run(tt.name, func(t *testing.T) {
+			if got := countLines([]byte(tt.content)); got != tt.want {
+				t.Errorf("countLines(%q) = %d, want %d", tt.content, got, tt.want)
+			}
+		})
+	}
+}
+
 // TestSplitLinesLongLine verifies the scanner gracefully handles very long lines
 // (up to the 1MB buffer limit set in C-05 fix).
 func TestSplitLinesLongLine(t *testing.T) {