diff --git a/Makefile b/Makefile
index b5695881..53a94c56 100644
--- a/Makefile
+++ b/Makefile
@@ -935,6 +935,9 @@ sizelint:
 test: all sizelint $(TESTS)
 	@WW=$(BIN)/ww BIN=$(BIN) sh test/run
 
+test-unit: all sizelint $(TESTS)
+	@WW=$(BIN)/ww BIN=$(BIN) UNIT=1 sh test/run
+
 install: all
 	mkdir -p $(PREFIX)/bin $(PREFIX)/lib
 	cp $(BIN)/ww $(PREFIX)/bin/
@@ -1080,4 +1083,4 @@ nocc:
 	@echo
 	@echo "NOCC OK: $(STAGE0)/* reproduces itself from source. cc not invoked."
 
-.PHONY: all cstage wwstage test sizelint install clean bootstrap nocc bootstrap-snapshot
+.PHONY: all cstage wwstage test test-unit sizelint install clean bootstrap nocc bootstrap-snapshot
diff --git a/test/run b/test/run
index 7a2499e3..4f547b02 100755
--- a/test/run
+++ b/test/run
@@ -46,15 +46,34 @@ trap 'rm -rf "$RESULTS"' EXIT
 
 JOBS=${JOBS:-$(nproc 2>/dev/null || echo 4)}
 
-# Emit (prefix, testpath) pairs NUL-separated. The prefix carries the
-# test's <NNN>_<name> stem so the collector iterates in lexical order,
-# yielding deterministic output even though workers interleave.
+# Phase 2 runs the wwstage byte-id gates (990-997) + 950_selfcheck
+# sequentially. ww_ww writes intermediates next to every traversed source
+# (task #15), so 991/992/994's reads of selfhost/cmd/<tool>/main.{s,o,
+# combined.ww} race against any concurrent driver-using test that emits
+# a sibling .combined.ww (task #16). Phase 1 stays parallel for unit
+# coverage; phase 2 sequential keeps the byte-id gates honest.
 for t in test/wcc/*.c; do
 	[ -f "$t" ] || continue
+	case ${t##*/} in
+	950_*|990_*|991_*|992_*|993_*|994_*|995_*|996_*|997_*) continue ;;
+	esac
 	name=${t##*/}; name=${name%.c}
 	printf '%s/%s\0%s\0' "$RESULTS" "$name" "$t"
 done | xargs -0 -n2 -P "$JOBS" "$0" --one || true
 
+# UNIT=1 (make test-unit) is the inner-loop short path: skip the wwstage
+# byte-id gates (990-997) and 950_selfcheck entirely. Pre-push uses
+# `make test` for the full suite.
+if [ "${UNIT:-0}" != "1" ]; then
+	for t in test/wcc/950_*.c test/wcc/990_*.c test/wcc/991_*.c \
+	         test/wcc/992_*.c test/wcc/993_*.c test/wcc/994_*.c \
+	         test/wcc/995_*.c test/wcc/996_*.c test/wcc/997_*.c; do
+		[ -f "$t" ] || continue
+		name=${t##*/}; name=${name%.c}
+		"$0" --one "$RESULTS/$name" "$t" || true
+	done
+fi
+
 fail=0
 ran=0
 for s in "$RESULTS"/*.status; do
diff --git a/test/wcc/995_self_rebuild.c b/test/wcc/995_self_rebuild.c
index ac0f072a..a188c15c 100644
--- a/test/wcc/995_self_rebuild.c
+++ b/test/wcc/995_self_rebuild.c
@@ -82,21 +82,18 @@ struct buildjob {
 /* Fork a child that runs the `ww_ww build` for this tool. The child
  * inherits no concurrent siblings — system() spawns a fresh /bin/sh -c.
  * stderr lands in <workdir>/build.err so concurrent builds don't merge
- * their diagnostics.
- *
- * The source is staged at <workdir>/build.ww before the build so the
- * driver writes its .combined.ww/.s/.o intermediates inside workdir —
- * not next to the canonical selfhost/cmd/<tool>/main.ww, where 991/992/
- * 994 read main.s as a fixture. Without staging, parallel test/run sees
- * 991 race against 995's wwdump-build mid-write of main.s. */
+ * their diagnostics. The driver leaves intermediates next to every
+ * source it traverses (selfhost/cmd/<tool>/main.s, etc.) — task #15 in
+ * the queue will let us redirect via `-o <outdir>`; until then, test/run
+ * serializes 990/995 after the parallel phase. */
 static int
 spawn_build(const char *bin, const char *cwd, struct buildjob *j)
 {
 	snprintf(j->workdir, sizeof j->workdir, "/tmp/wwsr_%d_%s",
 	    getpid(), j->tool);
-	char setup[4096];
-	snprintf(setup, sizeof setup, "rm -rf %s && mkdir -p %s && cp %s/%s %s/build.ww",
-	    j->workdir, j->workdir, cwd, j->src_rel, j->workdir);
+	char setup[256];
+	snprintf(setup, sizeof setup, "rm -rf %s && mkdir -p %s",
+	    j->workdir, j->workdir);
 	if (runwait(setup) != 0) return -1;
 
 	char cmd[4096];
@@ -104,15 +101,15 @@ spawn_build(const char *bin, const char *cwd, struct buildjob *j)
 		snprintf(cmd, sizeof cmd,
 		    "cd %s && timeout 180 %s/ww_ww build -I %s/%s "
 		    "-I %s/lib/ww -I %s/lib/ww/lex -I %s/lib/ww/parse "
-		    "-I %s/selfhost/cmd/wcc build.ww >/dev/null 2>%s/build.err",
+		    "-I %s/selfhost/cmd/wcc %s/%s >/dev/null 2>%s/build.err",
 		    j->workdir, bin, cwd, j->inc_local,
-		    cwd, cwd, cwd, cwd, j->workdir);
+		    cwd, cwd, cwd, cwd, cwd, j->src_rel, j->workdir);
 	} else {
 		snprintf(cmd, sizeof cmd,
 		    "cd %s && timeout 180 %s/ww_ww build "
 		    "-I %s/lib/ww -I %s/lib/ww/lex -I %s/lib/ww/parse "
-		    "-I %s/selfhost/cmd/wcc build.ww >/dev/null 2>%s/build.err",
-		    j->workdir, bin, cwd, cwd, cwd, cwd, j->workdir);
+		    "-I %s/selfhost/cmd/wcc %s/%s >/dev/null 2>%s/build.err",
+		    j->workdir, bin, cwd, cwd, cwd, cwd, cwd, j->src_rel, j->workdir);
 	}
 
 	pid_t p = fork();
@@ -152,7 +149,7 @@ collect_build(const char *bin, struct buildjob *j)
 		rc = -1;
 	} else {
 		char rebuilt[256], canonical[256];
-		snprintf(rebuilt,   sizeof rebuilt,   "%s/build", j->workdir);
+		snprintf(rebuilt,   sizeof rebuilt,   "%s/main", j->workdir);
 		snprintf(canonical, sizeof canonical, "%s/%s_ww", bin, j->tool);
 		rc = slurp_eq(rebuilt, canonical);
 		if (rc != 0) {