diff --git a/docs/debugger.md b/docs/debugger.md
index 06ae747..871ee6d 100644
--- a/docs/debugger.md
+++ b/docs/debugger.md
@@ -389,9 +389,12 @@ emit; sx provides the artifacts and a launch convenience, nothing more.
 
 ### Artifacts
 
-`sx build --emit-obj` / `--debug` writes the object (+DWARF) to a build
-dir and runs `dsymutil` to produce a `.dSYM`, so a debugger can load and
-symbolize it. Reuses the existing `emitObject` path.
+`sx build --emit-obj` keeps the DWARF-bearing object at its link-time path
+(`.sx-tmp/main.o`) instead of deleting it, and implies `-O0` (DWARF only emits
+at opt none/less). On **macOS** the linked binary's debug map resolves to that
+`.o`, so `lldb`/`gdb` run from the project root can step the binary directly; on
+**Linux** the DWARF is in the binary, so the `.o` isn't even needed. A portable
+`.dSYM` (via `dsymutil`) is only required for the on-device iOS rung (below).
 
 ### The verification ladder
 
@@ -400,8 +403,9 @@ Source-level stepping is verified manually/interactively (it needs
 provisioning profile — not a `run_examples.sh` test). Climb cheapest-first;
 the device run is the final sign-off:
 
-1. **macOS native** — `sx build --opt none` → `dsymutil` → drive
-   `lldb --batch` with a canned script: breakpoint on a sx function,
+1. **macOS native ✅** — `sx build --emit-obj` → drive `lldb --batch` (the
+   debug map resolves to the kept `.o`; no `dsymutil` needed locally).
+   Checked in as `tests/debug_stepping_smoke.sh`: breakpoint on a sx function,
    `run`, assert it stops at the right `.sx:line`, `next`/`stepi` advance,
    `bt` is source-mapped. The automatable rung (a checked-in smoke script).
 2. **iOS simulator** — bundle the `.app`, install to a booted simulator
@@ -438,8 +442,9 @@ a Mach-O debug map, never register JIT DWARF.
 | Niladic trace-push op + interned `Frame` table (runtime) | ✅ done — E3.3 slice 3a (`1b6cbc1`) |
 | Comptime resolver (`func_id, ir_offset` → location) | ✅ done — slice 3b |
 | Source snippet + `^` caret | ✅ done — slice 3c (line embedded in `Frame`) |
-| `--emit-obj` / `--debug` artifact plumbing | ⏳ planned — slice 3d |
-| Stepping verification ladder (macOS → sim → device) | ⏳ planned — slice 3e (capstone) |
+| `--emit-obj` artifact plumbing | ✅ done — slice 3d |
+| Stepping verification: macOS lldb | ✅ done — slice 3e rung 1 (`tests/debug_stepping_smoke.sh`) |
+| Stepping verification: iOS simulator → device | ⏳ planned — slice 3e rungs 2–3 (capstone) |
 | DWARF variable info (`DILocalVariable`, for `p x`) | ⏳ optional follow-on |
 
 The active plan and step breakdown live in `current/PLAN-ERR.md`
diff --git a/src/main.zig b/src/main.zig
index 83a1cf2..21bcc52 100644
--- a/src/main.zig
+++ b/src/main.zig
@@ -128,6 +128,8 @@ pub fn main(init: std.process.Init) !void {
             show_timing = true;
         } else if (std.mem.eql(u8, arg, "--cache")) {
             enable_cache = true;
+        } else if (std.mem.eql(u8, arg, "--emit-obj")) {
+            target_config.emit_obj = true;
         } else if (std.mem.startsWith(u8, arg, "-L")) {
             if (arg.len > 2) {
                 try lib_paths.append(allocator, arg[2..]);
@@ -183,6 +185,9 @@ pub fn main(init: std.process.Init) !void {
 
     if (std.mem.eql(u8, command, "build")) {
         target_config.is_aot = true;
+        // `--emit-obj` keeps a debuggable object; DWARF only emits at opt
+        // none/less, so default to -O0 unless the user set --opt explicitly.
+        if (target_config.emit_obj and !explicit_opt) target_config.opt_level = .none;
         const output_name = target_config.output_path orelse blk: {
             const base = deriveOutputName(path);
             if (target_config.isEmscripten()) {
@@ -390,6 +395,7 @@ fn printUsage() void {
         \\  --provisioning-profile <path>  .mobileprovision to embed (required for device)
         \\  --entitlements <path>  Entitlements plist (auto-extracted from profile if omitted)
         \\  --cache             Enable build caching
+        \\  --emit-obj          Keep the debuggable object (DWARF; implies -O0) for lldb/gdb
         \\  --time              Show compilation timing breakdown
         \\
     , .{});
@@ -570,8 +576,9 @@ fn compileWithTimer(allocator: std.mem.Allocator, io: std.Io, input_path: []cons
     const cache_obj = try cachePath(allocator, key, "o");
     const cache_bin = try cachePath(allocator, key, "bin");
 
-    // Level 1: Try cached binary (skip everything — no codegen, no link)
-    if (enable_cache) bin_cache: {
+    // Level 1: Try cached binary (skip everything — no codegen, no link).
+    // Skipped under --emit-obj, which needs the freshly-emitted object kept.
+    if (enable_cache and !target_config.emit_obj) bin_cache: {
         std.Io.Dir.copyFile(.cwd(), cache_bin, .cwd(), output_path, io, .{}) catch break :bin_cache;
         timer.record("cache");
         return;
@@ -776,14 +783,20 @@ fn compileWithTimer(allocator: std.mem.Allocator, io: std.Io, input_path: []cons
 
     std.debug.print("compiled: {s}\n", .{final_output});
 
-    // Clean up temp directory and all build artifacts
-    std.Io.Dir.deleteFile(.cwd(), io, obj_path) catch {};
+    // Clean up temp directory and all build artifacts. Under --emit-obj, keep
+    // the object (DWARF for lldb/gdb) at its link-time path — the binary's
+    // debug map resolves to it — and skip removing the temp dir.
     const shell_tmp = std.fmt.allocPrint(allocator, "{s}.shell.html", .{obj_path}) catch null;
     if (shell_tmp) |sp| std.Io.Dir.deleteFile(.cwd(), io, sp) catch {};
     for (c_obj_paths) |cop| {
         std.Io.Dir.deleteFile(.cwd(), io, cop) catch {};
     }
-    std.Io.Dir.deleteDir(.cwd(), io, tmp_dir) catch {};
+    if (target_config.emit_obj) {
+        std.debug.print("debug object kept: {s} (DWARF; run lldb/gdb from the project root)\n", .{obj_path});
+    } else {
+        std.Io.Dir.deleteFile(.cwd(), io, obj_path) catch {};
+        std.Io.Dir.deleteDir(.cwd(), io, tmp_dir) catch {};
+    }
 }
 
 fn runAOT(allocator: std.mem.Allocator, io: std.Io, input_path: []const u8, target_config: sx.target.TargetConfig, timer: *Timing, enable_cache: bool, stdlib_paths: []const []const u8) !void {
diff --git a/src/target.zig b/src/target.zig
index ab6b56b..0ec72b2 100644
--- a/src/target.zig
+++ b/src/target.zig
@@ -71,6 +71,13 @@ pub const TargetConfig = struct {
     /// `chdir` shouldn't run in JIT mode because it would mutate the host
     /// sx process's CWD.
     is_aot: bool = false,
+    /// Keep the DWARF-bearing object file after linking (`--emit-obj`) so a
+    /// debugger can step the binary: macOS resolves via the debug map → the
+    /// `.o`; Linux carries DWARF in the binary directly. Implies `-O0` unless
+    /// `--opt` is given explicitly (DWARF is only emitted at opt none/less).
+    /// The object is kept at `.sx-tmp/main.o` (its link-time path, so the
+    /// debug map resolves when lldb is run from the project root).
+    emit_obj: bool = false,
 
     pub const OptLevel = enum {
         none,
diff --git a/tests/debug_stepping_smoke.sh b/tests/debug_stepping_smoke.sh
new file mode 100755
index 0000000..003a10c
--- /dev/null
+++ b/tests/debug_stepping_smoke.sh
@@ -0,0 +1,54 @@
+#!/usr/bin/env bash
+# Debug-stepping smoke (ERR E3.0 slice 3e, rung 1: macOS native).
+#
+# Verifies the DWARF emitted by `sx build --emit-obj` actually drives
+# source-level stepping in lldb — the deep-debug half of the trace story.
+# NOT part of `run_examples.sh` (it needs `lldb`, and is macOS-specific via the
+# debug-map → kept `.o`). Run manually:  bash tests/debug_stepping_smoke.sh
+#
+# Exit 0 = lldb resolved a file:line breakpoint + a source-mapped backtrace.
+
+set -u
+ROOT_DIR="$(cd "$(dirname "$0")/.." && pwd)"
+SX="$ROOT_DIR/zig-out/bin/sx"
+TMP="$ROOT_DIR/.sx-tmp"
+SRC="$TMP/dbg_smoke.sx"
+BIN="$TMP/dbg_smoke"
+
+if ! command -v lldb >/dev/null 2>&1; then
+    echo "SKIP: lldb not found (macOS/Xcode tools required)"
+    exit 0
+fi
+
+mkdir -p "$TMP"
+cat > "$SRC" <<'EOF'
+add :: (a: s32, b: s32) -> s32 {
+    c := a + b;
+    return c;
+}
+main :: () -> s32 {
+    return add(40, 2);
+}
+EOF
+
+"$SX" build --emit-obj "$SRC" -o "$BIN" >/dev/null 2>&1 || { echo "FAIL: build"; exit 1; }
+
+# Breakpoint on the `return c;` line; expect lldb to resolve it + a backtrace
+# mapping both frames to dbg_smoke.sx.
+out=$(cd "$ROOT_DIR" && lldb --batch \
+    -o "breakpoint set --file dbg_smoke.sx --line 3" \
+    -o "run" -o "bt" -o "quit" "$BIN" 2>&1)
+
+rm -f "$SRC" "$BIN" "$TMP/main.o"
+
+fail=0
+echo "$out" | grep -q "dbg_smoke.sx:3" || { echo "FAIL: breakpoint did not resolve to dbg_smoke.sx:3"; fail=1; }
+echo "$out" | grep -q "add at dbg_smoke.sx:3" || { echo "FAIL: stopped frame not source-mapped"; fail=1; }
+echo "$out" | grep -q "main at dbg_smoke.sx:6" || { echo "FAIL: caller frame not source-mapped"; fail=1; }
+
+if [[ $fail -eq 0 ]]; then
+    echo "ok: lldb stepped sx source (breakpoint + backtrace resolved via DWARF)"
+    exit 0
+fi
+echo "--- lldb output ---"; echo "$out"
+exit 1
diff --git a/tests/expected/253-comptime-trace.exit b/tests/expected/253-comptime-trace.exit
index c227083..573541a 100644
--- a/tests/expected/253-comptime-trace.exit
+++ b/tests/expected/253-comptime-trace.exit
@@ -1 +1 @@
-0
\ No newline at end of file
+0