diff --git a/.kiro/specs/calculator/design.md b/.kiro/specs/calculator/design.md
index fdf75ea..bfe2961 100644
--- a/.kiro/specs/calculator/design.md
+++ b/.kiro/specs/calculator/design.md
@@ -614,7 +614,9 @@ Fields (navigable with Up/Down):
 4. HEX - editable hex (space-separated bytes: FF 00 AB CD)
 5. OCT - editable octal
 6. BIN - editable binary (nibble-separated: 1111 0000)
-7. Expression - full expression evaluator (& | ^ ~ << >> etc.)
+7. ASCII - shows printable ASCII interpretation of the bytes
+   (e.g., `.ELF` or `hell` - dots for non-printable)
+8. Expression - full expression evaluator (& | ^ ~ << >> etc.)
 
 Navigation:
 - Up/Down moves focus between fields
@@ -623,9 +625,25 @@ Navigation:
 - When bit grid is focused: left/right/up/down navigate, Space toggles
 - Tab cycles between Standard/Programmer mode
 - Ctrl-W cycles bit width (8/16/32/64)
+- Ctrl-E toggles endianness display (LE/BE)
 
-This replaces the current model where there's a passive display +
-separate input prompt at the bottom.
+Endianness:
+- Default: little-endian (matches x86-64)
+- Toggle with Ctrl-E
+- Affects: HEX byte display order, ASCII interpretation order
+- Does NOT affect: the underlying u64 value, DEC display, BIN display
+- Header shows current setting: `Endian: LE` or `Endian: BE`
+
+Text/ASCII literals:
+- Single-quoted strings in expressions: `'ELF'`, `'RIFF'`, `'a'`
+- Each character becomes one byte packed into the integer
+- Packing order follows endianness setting:
+  - BE: first char -> most significant used byte (`'AB'` = 0x4142)
+  - LE: first char -> least significant byte (`'AB'` = 0x4241)
+- Max chars = bit_width / 8 (1 for 8-bit, 2 for 16, 4 for 32, 8 for 64)
+- Non-ASCII characters are an error
+- Useful for: magic numbers (`'ELF'`), protocol tags, inspecting
+  how string bytes pack into integer registers
 
 ### 8.3 Struct Visualizer (sub-view of Programmer Mode)
 ```
diff --git a/.kiro/specs/calculator/requirements.md b/.kiro/specs/calculator/requirements.md
index d111a40..19345cb 100644
--- a/.kiro/specs/calculator/requirements.md
+++ b/.kiro/specs/calculator/requirements.md
@@ -17,17 +17,20 @@ A calculator application with three frontends (CLI, TUI, Android) sharing a comm
 - **FR-1.5**: Support constants: `pi`, `e`, `tau`.
 - **FR-1.6**: Support variable storage (Ans for last result, named variables A-F, X, Y, Z).
 - **FR-1.7**: Maintain calculation history with replay capability.
-- **FR-1.8**: Support implicit multiplication (e.g., `2pi`, `3(4+5)`).
+- **FR-1.8**: Commas accepted as digit separators in input (e.g., `1,000 * 2` = 2000). Spaces in hex literals treated as byte separators (e.g., `0xFF FF` = 0xFFFF).
 
 ### FR-2: Programmer Mode
 
 - **FR-2.1**: Accept input in decimal, hexadecimal (`0x`), octal (`0o`), and binary (`0b`) formats.
 - **FR-2.2**: Simultaneously display results in all four bases (dec, hex, oct, bin).
-- **FR-2.3**: Support configurable bit widths: 8, 16, 32, 64-bit.
+- **FR-2.3**: Support configurable bit widths: 8, 16, 32, 64, 128-bit.
 - **FR-2.4**: Support bitwise operators: AND (`&`), OR (`|`), XOR (`^`), NOT (`~`), left shift (`<<`), right shift (logical `>>>`), arithmetic right shift (`>>`), rotate left (`rol`), rotate right (`ror`).
 - **FR-2.5**: Display both signed (two's complement) and unsigned interpretations of the current value.
 - **FR-2.6**: Visualize the bit pattern as a grid (integer.exposed style) - bits individually addressable/toggleable in TUI and Android.
 - **FR-2.7**: Quick toggle between base display formats via dedicated keyboard shortcuts (TUI).
+- **FR-2.8**: Configurable endianness display (little-endian default, big-endian available). Affects byte order in HEX display and byte-level visualizations. Does not change the underlying value - purely a display/interpretation toggle.
+- **FR-2.9**: ASCII/text literal input via single-quoted strings: `'hello'` packs ASCII bytes into the integer value. First character occupies the most significant used byte (big-endian packing when display is BE) or least significant byte (when display is LE). Useful for examining magic numbers, file signatures, protocol headers.
+- **FR-2.10**: ASCII interpretation display - when the current value contains printable ASCII bytes, show the text representation alongside the numeric bases (e.g., `ASCII: "ELF."` or `ASCII: ..lf` with dots for non-printable bytes).
 
 ### FR-3: Struct Layout Visualizer (Programmer Mode Extension)
 
@@ -170,7 +173,7 @@ A calculator application with three frontends (CLI, TUI, Android) sharing a comm
 
 - Decimal numbers must use comma grouping for display (e.g., `4,294,967,295`).
 - Scientific notation only when value exceeds 15 significant digits or absolute value > 10^15 / < 10^-15. Never jump to scientific notation for values that fit in a readable decimal.
-- Programmer mode hex values display with underscore grouping per 16-bit word (e.g., `0xFFFF_FFFF`).
+- Programmer mode hex values display with space-separated bytes (e.g., `FF FF FF FF`).
 - Programmer mode binary values display grouped by nibble with spaces (e.g., `1111 1111`).
 - All frontends must distinguish between "display format" (with separators) and "clipboard format" (raw, no separators).
 - Android: long-press any result to copy the raw (no-separator) value to clipboard.
diff --git a/.kiro/specs/calculator/tasks.md b/.kiro/specs/calculator/tasks.md
index f894d76..c130ee1 100644
--- a/.kiro/specs/calculator/tasks.md
+++ b/.kiro/specs/calculator/tasks.md
@@ -245,15 +245,28 @@ Subcommands deferred until Phase 2 engine work is done.
 - Replace passive display + input prompt with navigable editable fields
 - Each base representation (DEC signed, DEC unsigned, HEX, OCT, BIN) is an
   editable text field
-- Up/Down arrows move focus between fields (bit grid, dec, hex, oct, bin, expression)
+- Up/Down arrows move focus between fields (bit grid, dec, hex, oct, bin, ascii, expression)
 - Typing in any field parses the value in that base and updates all other fields
 - Bit grid is one of the navigable fields (not a separate mode toggled with Ctrl-B)
 - Expression field at the bottom for full programmer-mode expressions
 - Remove Ctrl-B keybinding (field navigation replaces it)
 - HEX field accepts/displays space-separated bytes (e.g. "FF 00 AB CD")
+- Add ASCII display field showing printable byte interpretation (dots for non-printable)
+- Add Ctrl-E to toggle endianness (affects HEX byte order and ASCII interpretation)
+- Add endianness indicator to the header line (`Endian: LE` / `Endian: BE`)
 - Verify: can arrow between fields, edit hex and see decimal update, toggle bits
   and see all fields update, evaluate expression and see result in all fields
 
+### Task 5.3.2: Implement text/ASCII literal parsing [DONE]
+- Added single-quote string token to the tokenizer (`'hello'`)
+- Parser recognizes string literals as a new AST node type
+- Programmer evaluator packs ASCII bytes into u128 value (BE packing: first char -> MSB)
+- Standard evaluator also supports it (returns f64)
+- Enforces max length = bit_width / 8 (Overflow error if too many chars)
+- Rejects non-ASCII characters (> 0x7F)
+- 128-bit width also added (u64 -> u128 throughout engine)
+- Verify: `'A'` = 0x41, `'ELF'` = 0x454C46, overflow on too-long strings
+
 ### Task 5.4: Implement programmer mode TUI view - bit grid [DONE - merged into 5.3]
 - Bit grid widget: 8 bits per group, groups separated by space
 - Arrow key navigation (left/right moves cursor across bits, up/down between rows for 32/64-bit)
diff --git a/engine/src/ast.zig b/engine/src/ast.zig
index abdd298..e8ad761 100644
--- a/engine/src/ast.zig
+++ b/engine/src/ast.zig
@@ -9,6 +9,7 @@ const Base = types.Base;
 /// A single expression node.
 pub const Expr = union(enum) {
     number: Number,
+    string_literal: []const u8, // ASCII bytes to pack into integer (without quotes)
     unary: Unary,
     binary: Binary,
     call: Call,
diff --git a/engine/src/evaluator.zig b/engine/src/evaluator.zig
index 5766be6..c5acfc1 100644
--- a/engine/src/evaluator.zig
+++ b/engine/src/evaluator.zig
@@ -64,6 +64,15 @@ pub const Environment = struct {
 pub fn evaluate(env: *Environment, expr: *const Expr) CalcError!f64 {
     switch (expr.*) {
         .number => |n| return n.float_value,
+        .string_literal => |text| {
+            // Pack ASCII bytes into integer (same as programmer mode, BE packing)
+            var result: u128 = 0;
+            for (text) |byte| {
+                if (byte > 0x7F) return CalcError.InvalidNumber;
+                result = (result << 8) | byte;
+            }
+            return @floatFromInt(result);
+        },
         .variable => |name| {
             return env.getVar(name) orelse return CalcError.UnknownVariable;
         },
@@ -80,7 +89,8 @@ pub fn evaluate(env: *Environment, expr: *const Expr) CalcError!f64 {
                     // In standard mode, bitwise not doesn't really make sense,
                     // but we'll compute it on the integer representation
                     const int_val: u64 = @bitCast(@as(i64, @intFromFloat(operand)));
-                    const result = ~int_val & env.programmer_config.bit_width.mask();
+                    const mask_val: u64 = @truncate(env.programmer_config.bit_width.mask());
+                    const result = ~int_val & mask_val;
                     return @floatFromInt(@as(i64, @bitCast(result)));
                 },
             };
@@ -335,13 +345,13 @@ test "eval tau constant" {
     try testing.expectApproxEqAbs(math.tau, result, 1e-10);
 }
 
-test "eval implicit mul: 2pi" {
-    const result = try testEval("2pi");
+test "eval 2*pi" {
+    const result = try testEval("2*pi");
     try testing.expectApproxEqAbs(2.0 * math.pi, result, 1e-10);
 }
 
-test "eval implicit mul: 3(4+5)" {
-    const result = try testEval("3(4+5)");
+test "eval 3*(4+5)" {
+    const result = try testEval("3*(4+5)");
     try testing.expectEqual(@as(f64, 27.0), result);
 }
 
diff --git a/engine/src/formatter.zig b/engine/src/formatter.zig
index 394d3b5..30f0a87 100644
--- a/engine/src/formatter.zig
+++ b/engine/src/formatter.zig
@@ -29,7 +29,7 @@ pub fn formatFloat(buf: []u8, value: f64) FormattedValue {
 
     if (is_integer and @abs(value) < 1e15) {
         // Format as integer with commas
-        const int_val: i64 = @intFromFloat(value);
+        const int_val: i128 = @intFromFloat(value);
         const raw_len = writeSignedInt(buf, int_val);
         const raw = buf[0..raw_len];
 
@@ -57,7 +57,7 @@ pub fn formatFloat(buf: []u8, value: f64) FormattedValue {
 /// Format an integer for programmer mode hex display.
 /// Display: "FF FF FF FF" (space per byte)
 /// Raw: "0xFFFFFFFF" (no separators, with prefix)
-pub fn formatHex(buf: []u8, value: u64, bit_width: BitWidth) FormattedValue {
+pub fn formatHex(buf: []u8, value: u128, bit_width: BitWidth) FormattedValue {
     const width = bit_width.bits();
     const hex_digits: usize = @as(usize, width) / 4;
 
@@ -67,7 +67,7 @@ pub fn formatHex(buf: []u8, value: u64, bit_width: BitWidth) FormattedValue {
     var pos: usize = 2;
     var i: usize = 0;
     while (i < hex_digits) : (i += 1) {
-        const shift_amt: u6 = @intCast((hex_digits - 1 - i) * 4);
+        const shift_amt: u7 = @intCast((hex_digits - 1 - i) * 4);
         const nibble: u4 = @intCast((value >> shift_amt) & 0xF);
         buf[pos] = hexDigit(nibble);
         pos += 1;
@@ -82,7 +82,7 @@ pub fn formatHex(buf: []u8, value: u64, bit_width: BitWidth) FormattedValue {
             buf[pos] = ' ';
             pos += 1;
         }
-        const shift_amt: u6 = @intCast((hex_digits - 1 - i) * 4);
+        const shift_amt: u7 = @intCast((hex_digits - 1 - i) * 4);
         const nibble: u4 = @intCast((value >> shift_amt) & 0xF);
         buf[pos] = hexDigit(nibble);
         pos += 1;
@@ -95,7 +95,7 @@ pub fn formatHex(buf: []u8, value: u64, bit_width: BitWidth) FormattedValue {
 /// Format an integer for programmer mode binary display.
 /// Display: "1111 0000 1010 1100" (space per nibble)
 /// Raw: "0b1111000010101100" (no separators)
-pub fn formatBinary(buf: []u8, value: u64, bit_width: BitWidth) FormattedValue {
+pub fn formatBinary(buf: []u8, value: u128, bit_width: BitWidth) FormattedValue {
     const width: usize = bit_width.bits();
 
     // Write raw: "0b" + binary digits
@@ -104,7 +104,7 @@ pub fn formatBinary(buf: []u8, value: u64, bit_width: BitWidth) FormattedValue {
     var pos: usize = 2;
     var i: usize = 0;
     while (i < width) : (i += 1) {
-        const shift_amt: u6 = @intCast(width - 1 - i);
+        const shift_amt: u7 = @intCast(width - 1 - i);
         const bit: u8 = @intCast((value >> shift_amt) & 1);
         buf[pos] = '0' + bit;
         pos += 1;
@@ -119,7 +119,7 @@ pub fn formatBinary(buf: []u8, value: u64, bit_width: BitWidth) FormattedValue {
             buf[pos] = ' ';
             pos += 1;
         }
-        const shift_amt: u6 = @intCast(width - 1 - i);
+        const shift_amt: u7 = @intCast(width - 1 - i);
         const bit: u8 = @intCast((value >> shift_amt) & 1);
         buf[pos] = '0' + bit;
         pos += 1;
@@ -132,7 +132,7 @@ pub fn formatBinary(buf: []u8, value: u64, bit_width: BitWidth) FormattedValue {
 /// Format an integer for programmer mode octal display.
 /// Display: "0o37_777_777_777" (underscore per 3-digit group)
 /// Raw: "0o37777777777" (no separators)
-pub fn formatOctal(buf: []u8, value: u64, bit_width: BitWidth) FormattedValue {
+pub fn formatOctal(buf: []u8, value: u128, bit_width: BitWidth) FormattedValue {
     _ = bit_width; // Octal display shows all significant digits
 
     // Write raw octal
@@ -147,7 +147,7 @@ pub fn formatOctal(buf: []u8, value: u64, bit_width: BitWidth) FormattedValue {
     }
 
     // Calculate octal digits
-    var digits: [22]u8 = undefined; // max 22 octal digits for u64
+    var digits: [43]u8 = undefined; // max 43 octal digits for u128
     var digit_count: usize = 0;
     var v = value;
     while (v > 0) : (v /= 8) {
@@ -198,10 +198,10 @@ pub fn formatOctal(buf: []u8, value: u64, bit_width: BitWidth) FormattedValue {
     while (j > 0) {
         j -= 1;
         if (digits_written > 0 and digits_written == first_group) {
-            buf[pos] = '_';
+            buf[pos] = ' ';
             pos += 1;
         } else if (digits_written > first_group and (digits_written - first_group) % 3 == 0) {
-            buf[pos] = '_';
+            buf[pos] = ' ';
             pos += 1;
         }
         buf[pos] = '0' + digits[j];
@@ -216,7 +216,7 @@ pub fn formatOctal(buf: []u8, value: u64, bit_width: BitWidth) FormattedValue {
 /// Format an unsigned integer as decimal for programmer mode.
 /// Display: "4,294,967,295" (comma-separated)
 /// Raw: "4294967295" (no separators)
-pub fn formatDecimalUnsigned(buf: []u8, value: u64) FormattedValue {
+pub fn formatDecimalUnsigned(buf: []u8, value: u128) FormattedValue {
     const raw_len = writeUnsignedInt(buf, value);
     const raw = buf[0..raw_len];
 
@@ -230,7 +230,7 @@ pub fn formatDecimalUnsigned(buf: []u8, value: u64) FormattedValue {
 /// Format a signed integer as decimal for programmer mode.
 /// Display: "-1" or "4,294,967,295"
 /// Raw: same without commas
-pub fn formatDecimalSigned(buf: []u8, value: i64) FormattedValue {
+pub fn formatDecimalSigned(buf: []u8, value: i128) FormattedValue {
     const raw_len = writeSignedInt(buf, value);
     const raw = buf[0..raw_len];
 
@@ -248,12 +248,12 @@ fn hexDigit(nibble: u4) u8 {
     return 'A' + @as(u8, nibble) - 10;
 }
 
-fn writeUnsignedInt(buf: []u8, value: u64) usize {
+fn writeUnsignedInt(buf: []u8, value: u128) usize {
     if (value == 0) {
         buf[0] = '0';
         return 1;
     }
-    var digits: [20]u8 = undefined;
+    var digits: [39]u8 = undefined;
     var count: usize = 0;
     var v = value;
     while (v > 0) : (v /= 10) {
@@ -270,21 +270,21 @@ fn writeUnsignedInt(buf: []u8, value: u64) usize {
     return pos;
 }
 
-fn writeSignedInt(buf: []u8, value: i64) usize {
+fn writeSignedInt(buf: []u8, value: i128) usize {
     if (value < 0) {
         buf[0] = '-';
-        const abs_val: u64 = @intCast(-value);
+        const abs_val: u128 = @intCast(-value);
         return 1 + writeUnsignedInt(buf[1..], abs_val);
     }
     return writeUnsignedInt(buf, @intCast(value));
 }
 
-fn writeUnsignedWithCommas(buf: []u8, value: u64) usize {
+fn writeUnsignedWithCommas(buf: []u8, value: u128) usize {
     if (value == 0) {
         buf[0] = '0';
         return 1;
     }
-    var digits: [20]u8 = undefined;
+    var digits: [39]u8 = undefined;
     var count: usize = 0;
     var v = value;
     while (v > 0) : (v /= 10) {
@@ -308,10 +308,10 @@ fn writeUnsignedWithCommas(buf: []u8, value: u64) usize {
     return pos;
 }
 
-fn writeDecimalWithCommas(buf: []u8, value: i64) usize {
+fn writeDecimalWithCommas(buf: []u8, value: i128) usize {
     if (value < 0) {
         buf[0] = '-';
-        const abs_val: u64 = @intCast(-value);
+        const abs_val: u128 = @intCast(-value);
         return 1 + writeUnsignedWithCommas(buf[1..], abs_val);
     }
     return writeUnsignedWithCommas(buf, @intCast(value));
@@ -423,7 +423,7 @@ test "formatOctal: large with grouping" {
     var buf: [256]u8 = undefined;
     const result = formatOctal(&buf, 0xFFFF_FFFF, .bits32);
     // 0xFFFFFFFF = 37777777777 octal (11 digits)
-    try testing.expectEqualStrings("0o37_777_777_777", result.display);
+    try testing.expectEqualStrings("0o37 777 777 777", result.display);
     try testing.expectEqualStrings("0o37777777777", result.raw);
 }
 
diff --git a/engine/src/parser.zig b/engine/src/parser.zig
index 66750b4..cea4f70 100644
--- a/engine/src/parser.zig
+++ b/engine/src/parser.zig
@@ -80,20 +80,6 @@ pub const Parser = struct {
         var left = try self.parsePrefix();
 
         while (true) {
-            // Check for implicit multiplication first:
-            // If current token starts a value and isn't an operator, insert mul.
-            if (self.isImplicitMul()) {
-                const mul_prec = Prec.multiplicative;
-                if (@intFromEnum(mul_prec) <= @intFromEnum(min_prec)) break;
-                const right = try self.parseExpr(mul_prec);
-                left = try self.makeNode(.{ .binary = .{
-                    .op = .mul,
-                    .left = left,
-                    .right = right,
-                } });
-                continue;
-            }
-
             const prec = self.infixPrecedence(self.current.kind);
             if (@intFromEnum(prec) <= @intFromEnum(min_prec)) break;
 
@@ -117,6 +103,14 @@ pub const Parser = struct {
                     .base = num.base,
                 } });
             },
+            .string_literal => {
+                self.advance();
+                const text = tok.text(self.source);
+                // Strip quotes: 'abc' -> abc
+                if (text.len < 2) return CalcError.InvalidNumber;
+                const content = text[1 .. text.len - 1];
+                return self.makeNode(.{ .string_literal = content });
+            },
             .identifier => {
                 self.advance();
                 const name = tok.text(self.source);
@@ -249,31 +243,6 @@ pub const Parser = struct {
         } });
     }
 
-    /// Determine if the current position represents an implicit multiplication.
-    /// This happens when the current token could start a new value expression
-    /// and the previous token ended a value expression, with no operator between.
-    fn isImplicitMul(self: *Parser) bool {
-        // Implicit mul only happens when infixPrecedence returns .none
-        // (meaning the current token is NOT a recognized infix operator)
-        // but IS the start of a value expression.
-        if (self.infixPrecedence(self.current.kind) != .none) return false;
-
-        return switch (self.current.kind) {
-            .number, .left_paren => true,
-            .identifier => {
-                // Don't treat keyword operators as implicit mul
-                const name = self.current.text(self.source);
-                if (std.mem.eql(u8, name, "rol") or std.mem.eql(u8, name, "ror") or
-                    std.mem.eql(u8, name, "to"))
-                {
-                    return false;
-                }
-                return true;
-            },
-            else => false,
-        };
-    }
-
     /// Get the infix precedence of a token kind.
     fn infixPrecedence(self: *Parser, kind: TokenKind) Prec {
         return switch (kind) {
@@ -360,6 +329,7 @@ fn testParseArena(source: []const u8, mode: Mode) CalcError!*Expr {
 fn freeExpr(allocator: Allocator, expr: *Expr) void {
     switch (expr.*) {
         .number => {},
+        .string_literal => {},
         .variable => {},
         .unary => |u| freeExpr(allocator, u.operand),
         .binary => |b| {
@@ -465,31 +435,22 @@ test "parse assignment" {
     try testing.expectEqual(@as(f64, 42.0), expr.assignment.value.number.float_value);
 }
 
-test "parse implicit mul: number identifier" {
-    // 2pi should parse as 2 * pi
-    const expr = try testParse("2pi", .standard);
-    defer freeExpr(testing.allocator, expr);
-    try testing.expectEqual(BinaryOp.mul, expr.binary.op);
-    try testing.expectEqual(@as(f64, 2.0), expr.binary.left.number.float_value);
-    try testing.expectEqualStrings("pi", expr.binary.right.variable);
+test "parse adjacent number and identifier is an error (no implicit mul)" {
+    defer _ = test_arena_instance.reset(.retain_capacity);
+    const result = testParseArena("2pi", .standard);
+    try testing.expectError(CalcError.UnexpectedToken, result);
 }
 
-test "parse implicit mul: number paren" {
-    // 3(4+5) should parse as 3 * (4+5)
-    const expr = try testParse("3(4+5)", .standard);
-    defer freeExpr(testing.allocator, expr);
-    try testing.expectEqual(BinaryOp.mul, expr.binary.op);
-    try testing.expectEqual(@as(f64, 3.0), expr.binary.left.number.float_value);
-    try testing.expectEqual(BinaryOp.add, expr.binary.right.binary.op);
+test "parse adjacent number and paren is an error (no implicit mul)" {
+    defer _ = test_arena_instance.reset(.retain_capacity);
+    const result = testParseArena("3(4+5)", .standard);
+    try testing.expectError(CalcError.UnexpectedToken, result);
 }
 
-test "parse implicit mul: paren paren" {
-    // (2)(3) should parse as (2) * (3)
-    const expr = try testParse("(2)(3)", .standard);
-    defer freeExpr(testing.allocator, expr);
-    try testing.expectEqual(BinaryOp.mul, expr.binary.op);
-    try testing.expectEqual(@as(f64, 2.0), expr.binary.left.number.float_value);
-    try testing.expectEqual(@as(f64, 3.0), expr.binary.right.number.float_value);
+test "parse adjacent paren paren is an error (no implicit mul)" {
+    defer _ = test_arena_instance.reset(.retain_capacity);
+    const result = testParseArena("(2)(3)", .standard);
+    try testing.expectError(CalcError.UnexpectedToken, result);
 }
 
 test "parse caret as XOR in programmer mode" {
diff --git a/engine/src/programmer.zig b/engine/src/programmer.zig
index 8943136..954c846 100644
--- a/engine/src/programmer.zig
+++ b/engine/src/programmer.zig
@@ -1,6 +1,6 @@
 //! Programmer mode evaluator for Tally.
 //!
-//! All operations use exact integer arithmetic (u64 storage), with results
+//! All operations use exact integer arithmetic (u128 storage), with results
 //! masked to the configured bit width. No floating-point involved.
 //! Produces Integer values with signed/unsigned interpretation.
 
@@ -26,8 +26,8 @@ pub fn evalProgrammer(config: ProgrammerConfig, expr: *const Expr) CalcError!Int
     };
 }
 
-/// Recursively evaluate an expression to a raw u64.
-fn evalExpr(config: ProgrammerConfig, expr: *const Expr) CalcError!u64 {
+/// Recursively evaluate an expression to a raw u128.
+fn evalExpr(config: ProgrammerConfig, expr: *const Expr) CalcError!u128 {
     switch (expr.*) {
         .number => |n| {
             if (n.int_value) |int_val| {
@@ -35,15 +35,25 @@ fn evalExpr(config: ProgrammerConfig, expr: *const Expr) CalcError!u64 {
             }
             // Float in programmer mode: truncate to integer
             if (n.float_value < 0) {
-                const neg: i64 = @intFromFloat(n.float_value);
-                return @as(u64, @bitCast(neg)) & config.bit_width.mask();
+                const neg: i128 = @intFromFloat(n.float_value);
+                return @as(u128, @bitCast(neg)) & config.bit_width.mask();
             }
-            const val: u64 = @intFromFloat(n.float_value);
+            const val: u128 = @intFromFloat(n.float_value);
             return val & config.bit_width.mask();
         },
+        .string_literal => |text| {
+            // Pack ASCII bytes into integer.
+            // Big-endian packing: first char -> most significant used byte.
+            const max_bytes = @as(usize, config.bit_width.bits()) / 8;
+            if (text.len > max_bytes) return CalcError.Overflow;
+            var result: u128 = 0;
+            for (text) |byte| {
+                if (byte > 0x7F) return CalcError.InvalidNumber;
+                result = (result << 8) | byte;
+            }
+            return result & config.bit_width.mask();
+        },
         .variable => {
-            // Programmer mode doesn't support named variables (yet)
-            // Could add register-style variables later
             return CalcError.UnknownVariable;
         },
         .assignment => {
@@ -72,12 +82,12 @@ fn evalExpr(config: ProgrammerConfig, expr: *const Expr) CalcError!u64 {
     }
 }
 
-/// Evaluate a binary operation on two u64 values, masked to bit width.
-fn evalBinaryOp(config: ProgrammerConfig, op: BinaryOp, left: u64, right: u64) CalcError!u64 {
+/// Evaluate a binary operation on two u128 values, masked to bit width.
+fn evalBinaryOp(config: ProgrammerConfig, op: BinaryOp, left: u128, right: u128) CalcError!u128 {
     const mask = config.bit_width.mask();
     const width = config.bit_width.bits();
 
-    const result: u64 = switch (op) {
+    const result: u128 = switch (op) {
         .add => (left +% right) & mask,
         .sub => (left -% right) & mask,
         .mul => (left *% right) & mask,
@@ -93,7 +103,7 @@ fn evalBinaryOp(config: ProgrammerConfig, op: BinaryOp, left: u64, right: u64) C
             // Integer exponentiation
             var base = left;
             var exp = right;
-            var acc: u64 = 1;
+            var acc: u128 = 1;
             while (exp > 0) : (exp >>= 1) {
                 if (exp & 1 != 0) acc = (acc *% base) & mask;
                 base = (base *% base) & mask;
@@ -104,7 +114,7 @@ fn evalBinaryOp(config: ProgrammerConfig, op: BinaryOp, left: u64, right: u64) C
         .bit_or => left | right,
         .bit_xor => left ^ right,
         .shift_left => blk: {
-            const shift_amt: u6 = if (right >= width)
+            const shift_amt: u7 = if (right >= width)
                 @intCast(width - 1)
             else
                 @intCast(right);
@@ -112,38 +122,38 @@ fn evalBinaryOp(config: ProgrammerConfig, op: BinaryOp, left: u64, right: u64) C
         },
         .shift_right => blk: {
             // Arithmetic right shift: preserves sign bit
-            const shift_amt: u6 = if (right >= width)
+            const shift_amt: u7 = if (right >= width)
                 @intCast(width - 1)
             else
                 @intCast(right);
             // Sign-extend, shift, then mask
-            const sign_bit: u64 = @as(u64, 1) << @intCast(width - 1);
+            const sign_bit: u128 = @as(u128, 1) << @intCast(width - 1);
             if (left & sign_bit != 0) {
                 // Negative: fill with 1s from the top
                 const extended = left | ~mask;
-                const shifted: u64 = @bitCast(@as(i64, @bitCast(extended)) >> shift_amt);
+                const shifted: u128 = @bitCast(@as(i128, @bitCast(extended)) >> shift_amt);
                 break :blk shifted & mask;
             }
             break :blk (left >> shift_amt) & mask;
         },
         .shift_right_logical => blk: {
             // Logical right shift: always fills with 0s
-            const shift_amt: u6 = if (right >= width)
+            const shift_amt: u7 = if (right >= width)
                 @intCast(width - 1)
             else
                 @intCast(right);
             break :blk (left >> shift_amt) & mask;
         },
         .rotate_left => blk: {
-            const amt: u6 = @intCast(@mod(right, width));
+            const amt: u7 = @intCast(@mod(right, width));
             if (amt == 0) break :blk left;
-            const anti: u6 = @intCast(width - amt);
+            const anti: u7 = @intCast(width - amt);
             break :blk ((left << amt) | (left >> anti)) & mask;
         },
         .rotate_right => blk: {
-            const amt: u6 = @intCast(@mod(right, width));
+            const amt: u7 = @intCast(@mod(right, width));
             if (amt == 0) break :blk left;
-            const anti: u6 = @intCast(width - amt);
+            const anti: u7 = @intCast(width - amt);
             break :blk ((left >> amt) | (left << anti)) & mask;
         },
     };
@@ -174,45 +184,45 @@ fn testProgWith(source: []const u8, config: ProgrammerConfig) !Integer {
 
 test "prog: simple number" {
     const result = try testProg("42");
-    try testing.expectEqual(@as(u64, 42), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 42), result.unsignedValue());
 }
 
 test "prog: hex number" {
     const result = try testProg("0xFF");
-    try testing.expectEqual(@as(u64, 255), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 255), result.unsignedValue());
 }
 
 test "prog: binary number" {
     const result = try testProg("0b1010");
-    try testing.expectEqual(@as(u64, 10), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 10), result.unsignedValue());
 }
 
 test "prog: addition" {
     const result = try testProg("10 + 20");
-    try testing.expectEqual(@as(u64, 30), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 30), result.unsignedValue());
 }
 
 test "prog: subtraction wrapping" {
     // 5 - 10 in 8-bit unsigned wraps
     const result = try testProgWith("5 - 10", .{ .bit_width = .bits8 });
-    try testing.expectEqual(@as(u64, 251), result.unsignedValue()); // 256 - 5
-    try testing.expectEqual(@as(i64, -5), result.signedValue());
+    try testing.expectEqual(@as(u128, 251), result.unsignedValue()); // 256 - 5
+    try testing.expectEqual(@as(i128, -5), result.signedValue());
 }
 
 test "prog: multiplication" {
     const result = try testProg("6 * 7");
-    try testing.expectEqual(@as(u64, 42), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 42), result.unsignedValue());
 }
 
 test "prog: multiplication overflow 8-bit" {
     const result = try testProgWith("200 * 2", .{ .bit_width = .bits8 });
     // 400 & 0xFF = 144
-    try testing.expectEqual(@as(u64, 144), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 144), result.unsignedValue());
 }
 
 test "prog: division" {
     const result = try testProg("100 / 4");
-    try testing.expectEqual(@as(u64, 25), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 25), result.unsignedValue());
 }
 
 test "prog: division by zero" {
@@ -222,134 +232,134 @@ test "prog: division by zero" {
 
 test "prog: modulo" {
     const result = try testProg("10 % 3");
-    try testing.expectEqual(@as(u64, 1), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 1), result.unsignedValue());
 }
 
 test "prog: power" {
     const result = try testProg("2 ** 10");
-    try testing.expectEqual(@as(u64, 1024), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 1024), result.unsignedValue());
 }
 
 test "prog: bitwise AND" {
     const result = try testProg("0xFF & 0x0F");
-    try testing.expectEqual(@as(u64, 0x0F), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 0x0F), result.unsignedValue());
 }
 
 test "prog: bitwise OR" {
     const result = try testProg("0xF0 | 0x0F");
-    try testing.expectEqual(@as(u64, 0xFF), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 0xFF), result.unsignedValue());
 }
 
 test "prog: bitwise XOR" {
     const result = try testProg("0xFF ^ 0x0F");
-    try testing.expectEqual(@as(u64, 0xF0), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 0xF0), result.unsignedValue());
 }
 
 test "prog: bitwise NOT 8-bit" {
     const result = try testProgWith("~0x0F", .{ .bit_width = .bits8 });
-    try testing.expectEqual(@as(u64, 0xF0), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 0xF0), result.unsignedValue());
 }
 
 test "prog: bitwise NOT 16-bit" {
     const result = try testProgWith("~0x00FF", .{ .bit_width = .bits16 });
-    try testing.expectEqual(@as(u64, 0xFF00), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 0xFF00), result.unsignedValue());
 }
 
 test "prog: bitwise NOT 32-bit" {
     const result = try testProgWith("~0", .{ .bit_width = .bits32 });
-    try testing.expectEqual(@as(u64, 0xFFFF_FFFF), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 0xFFFF_FFFF), result.unsignedValue());
 }
 
 test "prog: shift left" {
     const result = try testProg("1 << 8");
-    try testing.expectEqual(@as(u64, 256), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 256), result.unsignedValue());
 }
 
 test "prog: shift left overflow 8-bit" {
     const result = try testProgWith("1 << 8", .{ .bit_width = .bits8 });
     // Shifting by width or more in 8-bit: shift_amt clamped to 7
-    try testing.expectEqual(@as(u64, 128), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 128), result.unsignedValue());
 }
 
 test "prog: logical shift right" {
     const result = try testProgWith("0x80 >>> 4", .{ .bit_width = .bits8 });
-    try testing.expectEqual(@as(u64, 0x08), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 0x08), result.unsignedValue());
 }
 
 test "prog: arithmetic shift right (sign bit preserved)" {
     // 0x80 in 8-bit is -128; >> 1 should give 0xC0 (-64)
     const result = try testProgWith("0x80 >> 1", .{ .bit_width = .bits8 });
-    try testing.expectEqual(@as(u64, 0xC0), result.unsignedValue());
-    try testing.expectEqual(@as(i64, -64), result.signedValue());
+    try testing.expectEqual(@as(u128, 0xC0), result.unsignedValue());
+    try testing.expectEqual(@as(i128, -64), result.signedValue());
 }
 
 test "prog: arithmetic shift right (positive)" {
     // 0x40 in 8-bit is positive; >> 1 should give 0x20
     const result = try testProgWith("0x40 >> 1", .{ .bit_width = .bits8 });
-    try testing.expectEqual(@as(u64, 0x20), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 0x20), result.unsignedValue());
 }
 
 test "prog: rotate left 8-bit" {
     // 0x81 rol 1 in 8-bit: bit 7 wraps to bit 0 -> 0x03
     const result = try testProgWith("0x81 rol 1", .{ .bit_width = .bits8 });
-    try testing.expectEqual(@as(u64, 0x03), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 0x03), result.unsignedValue());
 }
 
 test "prog: rotate right 8-bit" {
     // 0x81 ror 1 in 8-bit: bit 0 wraps to bit 7 -> 0xC0
     const result = try testProgWith("0x81 ror 1", .{ .bit_width = .bits8 });
-    try testing.expectEqual(@as(u64, 0xC0), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 0xC0), result.unsignedValue());
 }
 
 test "prog: negation two's complement" {
     const result = try testProgWith("-1", .{ .bit_width = .bits8 });
-    try testing.expectEqual(@as(u64, 0xFF), result.unsignedValue());
-    try testing.expectEqual(@as(i64, -1), result.signedValue());
+    try testing.expectEqual(@as(u128, 0xFF), result.unsignedValue());
+    try testing.expectEqual(@as(i128, -1), result.signedValue());
 }
 
 test "prog: negation 16-bit" {
     const result = try testProgWith("-42", .{ .bit_width = .bits16 });
-    try testing.expectEqual(@as(i64, -42), result.signedValue());
+    try testing.expectEqual(@as(i128, -42), result.signedValue());
 }
 
 test "prog: complex expression" {
     // (0xFF & 0x0F) | (1 << 4) = 0x0F | 0x10 = 0x1F
     const result = try testProg("(0xFF & 0x0F) | (1 << 4)");
-    try testing.expectEqual(@as(u64, 0x1F), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 0x1F), result.unsignedValue());
 }
 
 test "prog: precedence AND before OR" {
     // 0xF0 | 0xFF & 0x0F = 0xF0 | (0xFF & 0x0F) = 0xF0 | 0x0F = 0xFF
     const result = try testProg("0xF0 | 0xFF & 0x0F");
-    try testing.expectEqual(@as(u64, 0xFF), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 0xFF), result.unsignedValue());
 }
 
 test "prog: chained shifts" {
     const result = try testProg("1 << 4 << 2");
     // Left-associative: (1 << 4) << 2 = 16 << 2 = 64
-    try testing.expectEqual(@as(u64, 64), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 64), result.unsignedValue());
 }
 
 test "prog: mask applied to input" {
     // 0x1FF in 8-bit mode should be masked to 0xFF
     const result = try testProgWith("0x1FF", .{ .bit_width = .bits8 });
-    try testing.expectEqual(@as(u64, 0xFF), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 0xFF), result.unsignedValue());
 }
 
 test "prog: 32-bit operations" {
     const result = try testProgWith("0xFFFF_FFFF + 1", .{ .bit_width = .bits32 });
-    try testing.expectEqual(@as(u64, 0), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 0), result.unsignedValue());
 }
 
 test "prog: 64-bit max" {
     const result = try testProgWith("~0", .{ .bit_width = .bits64 });
-    try testing.expectEqual(@as(u64, 0xFFFF_FFFF_FFFF_FFFF), result.unsignedValue());
+    try testing.expectEqual(@as(u128, 0xFFFF_FFFF_FFFF_FFFF), result.unsignedValue());
 }
 
 test "prog: negative float input" {
     // -5.0 as a float in programmer mode should become two's complement
     const result = try testProgWith("-5", .{ .bit_width = .bits8 });
-    try testing.expectEqual(@as(i64, -5), result.signedValue());
+    try testing.expectEqual(@as(i128, -5), result.signedValue());
 }
 
 test "prog: variable reference errors" {
@@ -372,3 +382,31 @@ test "prog: function call errors" {
     const result = evalProgrammerString(arena.allocator(), "sin(1)", .{});
     try testing.expectError(CalcError.UnknownFunction, result);
 }
+
+test "prog: ASCII literal single char" {
+    const result = try testProg("'A'");
+    try testing.expectEqual(@as(u128, 0x41), result.unsignedValue());
+}
+
+test "prog: ASCII literal multi char" {
+    const result = try testProg("'ELF'");
+    try testing.expectEqual(@as(u128, 0x454C46), result.unsignedValue());
+}
+
+test "prog: ASCII literal full word" {
+    const result = try testProg("'ascii'");
+    try testing.expectEqual(@as(u128, 0x6173636969), result.unsignedValue());
+}
+
+test "prog: ASCII literal in expression" {
+    const result = try testProg("'A' | 0x20");
+    // 0x41 | 0x20 = 0x61 = 'a'
+    try testing.expectEqual(@as(u128, 0x61), result.unsignedValue());
+}
+
+test "prog: ASCII literal overflow 8-bit" {
+    var arena = std.heap.ArenaAllocator.init(std.heap.page_allocator);
+    defer _ = arena.deinit();
+    const result = evalProgrammerString(arena.allocator(), "'AB'", .{ .bit_width = .bits8 });
+    try testing.expectError(CalcError.Overflow, result);
+}
diff --git a/engine/src/tokenizer.zig b/engine/src/tokenizer.zig
index 1493967..717cc80 100644
--- a/engine/src/tokenizer.zig
+++ b/engine/src/tokenizer.zig
@@ -43,24 +43,8 @@ pub const TokenKind = enum {
     // Special
     eof,
     invalid,
-
-    /// Returns true if this token can be the last token before an implicit
-    /// multiplication (i.e. it produces a value).
-    pub fn isValueEnd(self: TokenKind) bool {
-        return switch (self) {
-            .number, .identifier, .right_paren => true,
-            else => false,
-        };
-    }
-
-    /// Returns true if this token can be the first token after an implicit
-    /// multiplication (i.e. it starts a value).
-    pub fn isValueStart(self: TokenKind) bool {
-        return switch (self) {
-            .number, .identifier, .left_paren, .tilde => true,
-            else => false,
-        };
-    }
+    // String literal (single-quoted, for ASCII byte packing in programmer mode)
+    string_literal,
 };
 
 pub const Token = struct {
@@ -93,7 +77,7 @@ pub fn parseNumber(token_text: []const u8) !NumberValue {
     var buf: [128]u8 = undefined;
     var buf_len: usize = 0;
     for (token_text) |c| {
-        if (c != '_' and c != ',') {
+        if (c != '_' and c != ',' and c != ' ') {
             if (buf_len >= buf.len) return error.InvalidNumber;
             buf[buf_len] = c;
             buf_len += 1;
@@ -151,51 +135,6 @@ pub fn parseNumber(token_text: []const u8) !NumberValue {
     return .{ .float = f, .int_value = null, .base = .decimal };
 }
 
-/// Token stream that inserts synthetic `star` tokens for implicit multiplication.
-pub const ImplicitMulStream = struct {
-    tokenizer: Tokenizer,
-    source: []const u8,
-    buffered: ?Token,
-    prev_kind: TokenKind,
-
-    pub fn init(source: []const u8, mode: Mode) ImplicitMulStream {
-        return .{
-            .tokenizer = Tokenizer.init(source, mode),
-            .source = source,
-            .buffered = null,
-            .prev_kind = .eof,
-        };
-    }
-
-    pub fn next(self: *ImplicitMulStream) Token {
-        // If we have a buffered token from a previous implicit mul detection
-        if (self.buffered) |buf| {
-            self.buffered = null;
-            self.prev_kind = buf.kind;
-            return buf;
-        }
-
-        const tok = self.tokenizer.next();
-
-        // Check for implicit multiplication:
-        // prev is value-end AND current is value-start
-        if (self.prev_kind.isValueEnd() and tok.kind.isValueStart()) {
-            // Buffer this token for the next call, emit star now
-            self.buffered = tok;
-            self.prev_kind = .star;
-            return .{ .kind = .star, .start = tok.start, .len = 0 };
-        }
-
-        self.prev_kind = tok.kind;
-        return tok;
-    }
-
-    /// Peek at the source text for a given token.
-    pub fn tokenText(self: *const ImplicitMulStream, tok: Token) []const u8 {
-        return tok.text(self.source);
-    }
-};
-
 // -- Raw Tokenizer --
 
 pub const Tokenizer = struct {
@@ -275,6 +214,7 @@ pub const Tokenizer = struct {
                 self.pos += 1;
                 return .{ .kind = .invalid, .start = start, .len = 1 };
             },
+            '\'' => return self.readStringLiteral(start),
             else => {
                 self.pos += 1;
                 return .{ .kind = .invalid, .start = start, .len = 1 };
@@ -303,7 +243,7 @@ pub const Tokenizer = struct {
             switch (next_ch) {
                 'x', 'X' => {
                     self.pos += 2;
-                    self.consumeDigits(isHexDigit);
+                    self.consumeHexDigits();
                     return .{ .kind = .number, .start = start, .len = self.pos - start };
                 },
                 'o', 'O' => {
@@ -381,6 +321,40 @@ pub const Tokenizer = struct {
         }
     }
 
+    /// Like consumeDigits but also treats spaces as byte separators.
+    /// Only used for hex literals (0x...) where "FF FF FF" is valid input.
+    fn consumeHexDigits(self: *Tokenizer) void {
+        while (self.pos < self.source.len) {
+            const ch = self.source[self.pos];
+            if (isHexDigit(ch)) {
+                self.pos += 1;
+            } else if (ch == '_') {
+                self.pos += 1;
+            } else if (ch == ' ') {
+                // Space in hex: only consume if followed by a hex digit
+                // (to avoid eating the space before an operator like "0xFF + 1")
+                if (self.pos + 1 < self.source.len and isHexDigit(self.source[self.pos + 1])) {
+                    self.pos += 1;
+                } else {
+                    break;
+                }
+            } else {
+                break;
+            }
+        }
+    }
+
+    fn readStringLiteral(self: *Tokenizer, start: usize) Token {
+        self.pos += 1; // consume opening quote
+        while (self.pos < self.source.len and self.source[self.pos] != '\'') {
+            self.pos += 1;
+        }
+        if (self.pos < self.source.len) {
+            self.pos += 1; // consume closing quote
+        }
+        return .{ .kind = .string_literal, .start = start, .len = self.pos - start };
+    }
+
     fn readIdentifier(self: *Tokenizer, start: usize) Token {
         while (self.pos < self.source.len) {
             const ch = self.source[self.pos];
@@ -618,59 +592,10 @@ test "parseNumber with commas" {
 
 // -- ImplicitMulStream tests --
 
-test "implicit mul: number followed by identifier (2pi)" {
-    var stream = ImplicitMulStream.init("2pi", .standard);
-    try testing.expectEqual(TokenKind.number, stream.next().kind);
-    try testing.expectEqual(TokenKind.star, stream.next().kind); // synthetic
-    try testing.expectEqual(TokenKind.identifier, stream.next().kind);
-    try testing.expectEqual(TokenKind.eof, stream.next().kind);
-}
-
-test "implicit mul: number followed by paren 3(4+5)" {
-    var stream = ImplicitMulStream.init("3(4+5)", .standard);
-    try testing.expectEqual(TokenKind.number, stream.next().kind);
-    try testing.expectEqual(TokenKind.star, stream.next().kind); // synthetic
-    try testing.expectEqual(TokenKind.left_paren, stream.next().kind);
-    try testing.expectEqual(TokenKind.number, stream.next().kind);
-    try testing.expectEqual(TokenKind.plus, stream.next().kind);
-    try testing.expectEqual(TokenKind.number, stream.next().kind);
-    try testing.expectEqual(TokenKind.right_paren, stream.next().kind);
-    try testing.expectEqual(TokenKind.eof, stream.next().kind);
-}
-
-test "implicit mul: paren followed by paren )()" {
-    var stream = ImplicitMulStream.init("(2)(3)", .standard);
-    try testing.expectEqual(TokenKind.left_paren, stream.next().kind);
-    try testing.expectEqual(TokenKind.number, stream.next().kind);
-    try testing.expectEqual(TokenKind.right_paren, stream.next().kind);
-    try testing.expectEqual(TokenKind.star, stream.next().kind); // synthetic
-    try testing.expectEqual(TokenKind.left_paren, stream.next().kind);
-    try testing.expectEqual(TokenKind.number, stream.next().kind);
-    try testing.expectEqual(TokenKind.right_paren, stream.next().kind);
-    try testing.expectEqual(TokenKind.eof, stream.next().kind);
-}
-
-test "implicit mul: identifier followed by paren (function call) does NOT insert star" {
-    // "sin(x)" - identifier followed by ( is a function call, not implicit mul.
-    // However, at the tokenizer level, identifier IS a value-end and ( IS a value-start,
-    // so the ImplicitMulStream WILL insert a star. The PARSER is responsible for
-    // recognizing "identifier * (" as a function call pattern (when the star is
-    // synthetic/zero-length) and handling it correctly.
-    //
-    // This test documents the actual stream behavior:
-    var stream = ImplicitMulStream.init("sin(x)", .standard);
-    try testing.expectEqual(TokenKind.identifier, stream.next().kind);
-    try testing.expectEqual(TokenKind.star, stream.next().kind); // synthetic
-    try testing.expectEqual(TokenKind.left_paren, stream.next().kind);
-    try testing.expectEqual(TokenKind.identifier, stream.next().kind);
-    try testing.expectEqual(TokenKind.right_paren, stream.next().kind);
-    try testing.expectEqual(TokenKind.eof, stream.next().kind);
-}
-
-test "no implicit mul: number + operator + number" {
-    var stream = ImplicitMulStream.init("2 + 3", .standard);
-    try testing.expectEqual(TokenKind.number, stream.next().kind);
-    try testing.expectEqual(TokenKind.plus, stream.next().kind);
-    try testing.expectEqual(TokenKind.number, stream.next().kind);
-    try testing.expectEqual(TokenKind.eof, stream.next().kind);
+test "no implicit mul: spaces are just whitespace" {
+    // Spaces between tokens don't create implicit multiplication
+    var tok = Tokenizer.init("2 3", .standard);
+    try testing.expectEqual(TokenKind.number, tok.next().kind);
+    try testing.expectEqual(TokenKind.number, tok.next().kind);
+    try testing.expectEqual(TokenKind.eof, tok.next().kind);
 }
diff --git a/engine/src/types.zig b/engine/src/types.zig
index b4f34bc..d930b1f 100644
--- a/engine/src/types.zig
+++ b/engine/src/types.zig
@@ -10,24 +10,26 @@ pub const Mode = enum {
 };
 
 /// Configurable integer bit width for programmer mode.
-pub const BitWidth = enum(u7) {
+pub const BitWidth = enum(u8) {
     bits8 = 8,
     bits16 = 16,
     bits32 = 32,
     bits64 = 64,
+    bits128 = 128,
 
     /// Returns the mask for this bit width (all bits set within width).
-    pub fn mask(self: BitWidth) u64 {
+    pub fn mask(self: BitWidth) u128 {
         return switch (self) {
             .bits8 => 0xFF,
             .bits16 => 0xFFFF,
             .bits32 => 0xFFFF_FFFF,
             .bits64 => 0xFFFF_FFFF_FFFF_FFFF,
+            .bits128 => 0xFFFF_FFFF_FFFF_FFFF_FFFF_FFFF_FFFF_FFFF,
         };
     }
 
     /// Returns the number of bits as a plain integer.
-    pub fn bits(self: BitWidth) u7 {
+    pub fn bits(self: BitWidth) u8 {
         return @intFromEnum(self);
     }
 };
@@ -53,22 +55,22 @@ pub const Base = enum {
 };
 
 /// An integer value in programmer mode.
-/// Raw bits stored in u64; interpretation depends on bit_width and signedness.
+/// Raw bits stored in u128; interpretation depends on bit_width and signedness.
 pub const Integer = struct {
-    raw: u64,
+    raw: u128,
     bit_width: BitWidth,
     signedness: Signedness,
 
     /// Apply the bit width mask, truncating to the configured width.
-    pub fn masked(self: Integer) u64 {
+    pub fn masked(self: Integer) u128 {
         return self.raw & self.bit_width.mask();
     }
 
     /// Interpret as a signed value (sign-extend from bit_width).
-    pub fn signedValue(self: Integer) i64 {
+    pub fn signedValue(self: Integer) i128 {
         const m = self.masked();
         const width = self.bit_width.bits();
-        const sign_bit: u64 = @as(u64, 1) << @intCast(width - 1);
+        const sign_bit: u128 = @as(u128, 1) << @intCast(width - 1);
         if (m & sign_bit != 0) {
             // Sign extend: fill upper bits with 1s
             const extension = ~self.bit_width.mask();
@@ -78,7 +80,7 @@ pub const Integer = struct {
     }
 
     /// Interpret as an unsigned value (just mask).
-    pub fn unsignedValue(self: Integer) u64 {
+    pub fn unsignedValue(self: Integer) u128 {
         return self.masked();
     }
 };
@@ -141,32 +143,33 @@ pub const ErrorInfo = struct {
 };
 
 test "BitWidth.mask" {
-    try std.testing.expectEqual(@as(u64, 0xFF), BitWidth.bits8.mask());
-    try std.testing.expectEqual(@as(u64, 0xFFFF), BitWidth.bits16.mask());
-    try std.testing.expectEqual(@as(u64, 0xFFFF_FFFF), BitWidth.bits32.mask());
-    try std.testing.expectEqual(@as(u64, 0xFFFF_FFFF_FFFF_FFFF), BitWidth.bits64.mask());
+    try std.testing.expectEqual(@as(u128, 0xFF), BitWidth.bits8.mask());
+    try std.testing.expectEqual(@as(u128, 0xFFFF), BitWidth.bits16.mask());
+    try std.testing.expectEqual(@as(u128, 0xFFFF_FFFF), BitWidth.bits32.mask());
+    try std.testing.expectEqual(@as(u128, 0xFFFF_FFFF_FFFF_FFFF), BitWidth.bits64.mask());
+    try std.testing.expectEqual(@as(u128, 0xFFFF_FFFF_FFFF_FFFF_FFFF_FFFF_FFFF_FFFF), BitWidth.bits128.mask());
 }
 
 test "Integer.signedValue" {
     // 0xFF in 8-bit signed = -1
     const i8_neg1 = Integer{ .raw = 0xFF, .bit_width = .bits8, .signedness = .signed };
-    try std.testing.expectEqual(@as(i64, -1), i8_neg1.signedValue());
+    try std.testing.expectEqual(@as(i128, -1), i8_neg1.signedValue());
 
     // 0x7F in 8-bit signed = 127
     const i8_127 = Integer{ .raw = 0x7F, .bit_width = .bits8, .signedness = .signed };
-    try std.testing.expectEqual(@as(i64, 127), i8_127.signedValue());
+    try std.testing.expectEqual(@as(i128, 127), i8_127.signedValue());
 
     // 0x80 in 8-bit signed = -128
     const i8_neg128 = Integer{ .raw = 0x80, .bit_width = .bits8, .signedness = .signed };
-    try std.testing.expectEqual(@as(i64, -128), i8_neg128.signedValue());
+    try std.testing.expectEqual(@as(i128, -128), i8_neg128.signedValue());
 
     // 0xFFFF_FFFF in 32-bit signed = -1
     const i32_neg1 = Integer{ .raw = 0xFFFF_FFFF, .bit_width = .bits32, .signedness = .signed };
-    try std.testing.expectEqual(@as(i64, -1), i32_neg1.signedValue());
+    try std.testing.expectEqual(@as(i128, -1), i32_neg1.signedValue());
 }
 
 test "Integer.unsignedValue masks correctly" {
     // Extra bits beyond width are masked off
     const val = Integer{ .raw = 0x1FF, .bit_width = .bits8, .signedness = .unsigned };
-    try std.testing.expectEqual(@as(u64, 0xFF), val.unsignedValue());
+    try std.testing.expectEqual(@as(u128, 0xFF), val.unsignedValue());
 }
diff --git a/src/main.zig b/src/main.zig
index 1172f75..5d36554 100644
--- a/src/main.zig
+++ b/src/main.zig
@@ -291,11 +291,11 @@ test "evaluate: function" {
     try testing.expectEqualStrings("2", result.output);
 }
 
-test "evaluate: implicit mul" {
+test "evaluate: explicit mul" {
     var arena = std.heap.ArenaAllocator.init(std.heap.page_allocator);
     defer _ = arena.deinit();
     var buf: [4096]u8 = undefined;
-    const result = evaluate(arena.allocator(), "3(4+5)", .standard, &buf);
+    const result = evaluate(arena.allocator(), "3*(4+5)", .standard, &buf);
     try testing.expect(!result.is_error);
     try testing.expectEqualStrings("27", result.output);
 }
diff --git a/src/tui.zig b/src/tui.zig
index d8c5982..734d2aa 100644
--- a/src/tui.zig
+++ b/src/tui.zig
@@ -29,9 +29,9 @@ pub const App = struct {
     saved_input: ?[]const u8,
     mode: Mode,
     // Programmer mode state
-    prog_value: u64,
+    prog_value: u128,
     prog_config: engine.types.ProgrammerConfig,
-    bit_cursor: u6,
+    bit_cursor: u7,
     bit_grid_active: bool,
 
     pub const HistoryEntry = struct {
@@ -114,7 +114,7 @@ pub const App = struct {
                 if (ans >= 0 and ans == @trunc(ans) and ans < 18446744073709551616.0) {
                     self.prog_value = @intFromFloat(ans);
                 } else if (ans < 0 and ans == @trunc(ans) and ans >= -9223372036854775808.0) {
-                    const signed: i64 = @intFromFloat(ans);
+                    const signed: i128 = @intFromFloat(ans);
                     self.prog_value = @bitCast(signed);
                 }
                 self.prog_value &= self.prog_config.bit_width.mask();
@@ -141,7 +141,8 @@ pub const App = struct {
                 .bits8 => .bits16,
                 .bits16 => .bits32,
                 .bits32 => .bits64,
-                .bits64 => .bits8,
+                .bits64 => .bits128,
+                .bits128 => .bits8,
             };
             self.prog_value &= self.prog_config.bit_width.mask();
             if (self.bit_cursor >= self.prog_config.bit_width.bits()) {
@@ -229,7 +230,7 @@ pub const App = struct {
             return true;
         }
         if (key.matches(' ', .{}) or key.matches(vaxis.Key.enter, .{})) {
-            self.prog_value ^= @as(u64, 1) << self.bit_cursor;
+            self.prog_value ^= @as(u128, 1) << self.bit_cursor;
             self.prog_value &= self.prog_config.bit_width.mask();
             return true;
         }
diff --git a/src/tui/programmer.zig b/src/tui/programmer.zig
index 92f56be..5bdfea4 100644
--- a/src/tui/programmer.zig
+++ b/src/tui/programmer.zig
@@ -72,9 +72,9 @@ pub fn drawProgrammerMode(app: *tui.App, surface: *vxfw.Surface, width: u16, hei
     draw.writeStr(surface, height -| 1, 1, status, .{ .fg = C.muted, .bg = C.bg });
 }
 
-fn drawBitGrid(app: *tui.App, surface: *vxfw.Surface, start_row: u16, val: u64, bw: engine.types.BitWidth) void {
+fn drawBitGrid(app: *tui.App, surface: *vxfw.Surface, start_row: u16, val: u128, bw: engine.types.BitWidth) void {
     const width_bits = bw.bits();
-    const bits_per_row: u7 = if (width_bits > 32) 32 else width_bits;
+    const bits_per_row: u8 = if (width_bits > 32) 32 else width_bits;
     const num_rows: u16 = (@as(u16, width_bits) + bits_per_row - 1) / bits_per_row;
 
     var row = start_row;