feat(core): INCLUDE, error overhaul, sf64 lane, WORDS ALL, .RS

WS-012 -- INCLUDE/INCLUDED:
- Injected source loader (core stays IO-free: CLI installs a
  filesystem reader, web leaves it unset -> defined error). Recursive
  include_file feeds files line-by-line through evaluate, so compile
  state and SEE capture span lines for free. Cycle detection, depth
  cap 16, paths relative to the including file, SOURCE-ID per nesting
  level, parent input restored on success/error/BYE.
- CLI file mode now runs through the include machinery: `wafer x.fth`
  gets file:line error context and a base dir for nested INCLUDEs.
- Unlocks the REMEMBER+INCLUDE reload loop.

WS-008 -- error reporting remainder:
- Errors inside included files carry `file.fth:12:` context
  (anyhow context chain; CLI prints {e:#}).
- describe_uncaught now returns typed WaferError::UncaughtThrow
  { code, message } -- display text unchanged, THROW code reachable
  via downcast for CLI/web consumers.
- compile_word emits a WASM name section; wasmtime trap backtraces
  name the faulting word and runtime_native prefixes "in <WORD>:".
  Batch/consolidated modules stay unnamed (no name plumbing there;
  boot primitives rarely trap).

WS-003 -- SwiftForth correctness lane:
- compare_all_programs_sf64 runs the program corpus with sf64 as
  oracle; whitespace-token comparison (sf64 prints numbers
  space-prefixed and echoes piped lines). 34/35 parity; dot-quote
  skipped (interpret-mode ." is a SwiftForth no-op). #[ignore]d like
  the gforth lane; `just compare-correctness` runs both.

WS-011 leftovers:
- WORDS ALL: grouped full view -- one section per wordlist (search
  order first), then internal words, each with counts. Backed by
  Dictionary::visible_entries (name, wid, internal); visible_words
  now derives from it.
- .RS / RDEPTH: return-stack introspection in boot.fth over a new
  RP@ primitive (IrOp::RpFetch); BEGIN/WHILE walk so the walk never
  touches the stack it prints. SPACES clamped per 6.1.2230.

549 unit + 11 compliance + 9(+2) comparison + 5 crypto + 1 bench
green; fmt/clippy clean; --no-default-features and wasm32 web checks
pass.

crates/core/src/codegen.rs

@@ -899,6 +899,15 @@ fn emit_op(f: &mut Function, op: &IrOp, ctx: &mut EmitCtx) {
                 .instruction(&Instruction::I32Store(MEM4));
         }
 
+        IrOp::RpFetch => {
+            // Push the current return-stack pointer onto the data stack.
+            // `$rsp` lives in a global (not cached), so no writeback needed.
+            dsp_dec(f);
+            f.instruction(&Instruction::LocalGet(CACHED_DSP_LOCAL))
+                .instruction(&Instruction::GlobalGet(RSP))
+                .instruction(&Instruction::I32Store(MEM4));
+        }
+
         // -- Compound operations -----------------------------------------------
         IrOp::TwoDup => {
             // ( a b -- a b a b )
@@ -1242,7 +1251,9 @@ fn is_promotable(ops: &[IrOp]) -> bool {
 fn is_promotable_body(ops: &[IrOp]) -> bool {
     for op in ops {
         match op {
-            IrOp::Call(_) | IrOp::TailCall(_) | IrOp::Execute | IrOp::SpFetch => return false,
+            IrOp::Call(_) | IrOp::TailCall(_) | IrOp::Execute | IrOp::SpFetch | IrOp::RpFetch => {
+                return false;
+            }
             IrOp::ToR | IrOp::FromR | IrOp::Exit => return false,
             IrOp::ForthLocalGet(_) | IrOp::ForthLocalSet(_) => return false,
             IrOp::ForthFLocalGet(_) | IrOp::ForthFLocalSet(_) => return false,
@@ -2306,6 +2317,9 @@ fn body_needs_return_stack(ops: &[IrOp]) -> bool {
         match op {
             IrOp::Call(_) | IrOp::TailCall(_) | IrOp::Execute => return true,
             IrOp::ToR | IrOp::FromR => return true,
+            // RP@ observes the return stack, so loop params must be there
+            // (otherwise inlined RDEPTH/.RS would report an empty stack).
+            IrOp::RpFetch => return true,
             // RFetch (I) is handled by loop locals in the fast path — not a problem.
             // LoopJ is also handled by loop locals.
             // Only explicit >R / R> / calls force the slow path.
@@ -2518,7 +2532,7 @@ fn count_forth_f_locals(ops: &[IrOp]) -> u32 {
 /// This is the JIT path: each word gets its own module that imports
 /// shared memory, globals, and function table from the host.
 pub fn compile_word(
-    _name: &str,
+    name: &str,
     body: &[IrOp],
     config: &CodegenConfig,
 ) -> WaferResult<CompiledModule> {
@@ -2685,6 +2699,16 @@ pub fn compile_word(
     code.function(&func);
     module.section(&code);
 
+    // -- Name section: carries the Forth word name into wasmtime trap
+    // backtraces (best-effort symbolication, WS-008).
+    let mut names = wasm_encoder::NameSection::new();
+    names.module(name);
+    let mut fn_names = wasm_encoder::NameMap::new();
+    fn_names.append(0, "emit");
+    fn_names.append(WORD_FUNC, name);
+    names.functions(&fn_names);
+    module.section(&names);
+
     let bytes = module.finish();
 
     // Validate