test(e2e): register e2e-live source require hook and surface cloud-onboard dcode identity failure

test/e2e/e2e-cloud-experimental/checks/04-deepagents-code-fresh-reonboard.sh

@@ -33,7 +33,11 @@ sandbox_exec() {
 }
 
 dcode_identity() {
-  openshell sandbox exec --name "$SANDBOX_NAME" -- dcode identity 2>&1
+  # Invoke dcode by absolute path: `openshell sandbox exec -- dcode ...` runs
+  # without a login shell, so /usr/local/bin is not on PATH and a bare `dcode`
+  # resolves to "command not found". The image installs the launcher at
+  # /usr/local/bin/dcode (see agents/langchain-deepagents-code/Dockerfile).
+  openshell sandbox exec --name "$SANDBOX_NAME" -- /usr/local/bin/dcode identity 2>&1
 }
 
 identity_field() {
@@ -148,7 +152,13 @@ fi
 [ -n "${COMPATIBLE_API_KEY:-}" ] || fail "COMPATIBLE_API_KEY is required"
 [ -x "$CLI" ] || fail "NemoClaw CLI is not executable at $CLI"
 
-identity_before="$(dcode_identity)" || fail "could not read initial dcode identity"
+if ! identity_before="$(dcode_identity)"; then
+  # Surface the captured stdout+stderr (dcode_identity redirects 2>&1) before
+  # failing. Without this the real reason `dcode identity` exits non-zero is
+  # discarded and CI/result.json only show the generic message with stdout "".
+  printf '%s: diagnostic: initial dcode identity output:\n%s\n' "$PREFIX" "${identity_before:-<no output captured>}"
+  fail "could not read initial dcode identity"
+fi
 model_a="$(identity_field "$identity_before" Model)"
 model_a="${model_a#openai:}"
 [ -n "$model_a" ] || fail "initial dcode identity did not report a model"
@@ -193,7 +203,10 @@ pass "same-name --fresh re-onboard crossed backup and restore boundaries"
 sandbox_list="$(openshell sandbox list 2>&1)" || fail "could not list sandbox after re-onboard"
 printf '%s\n' "$sandbox_list" | awk -v name="$SANDBOX_NAME" '$1 == name && /Ready/ { found = 1 } END { exit(found ? 0 : 1) }' || fail "same-name sandbox is not Ready after re-onboard"
 
-identity_after="$(dcode_identity)" || fail "could not read dcode identity after re-onboard"
+if ! identity_after="$(dcode_identity)"; then
+  printf '%s: diagnostic: dcode identity output after re-onboard:\n%s\n' "$PREFIX" "${identity_after:-<no output captured>}"
+  fail "could not read dcode identity after re-onboard"
+fi
 assert_identity "$identity_after" "$model_b" "fresh"
 printf '%s\n' "$identity_after" | grep -Fq "$model_a" && fail "fresh identity still contains model A"
 pass "live dcode identity reports model B"

vitest.config.ts

@@ -140,6 +140,13 @@ export default defineConfig({
         test: {
           name: "e2e-live",
           alias: canonicalOpenShellPolicyAlias,
+          // Register the typed-source require hook in the worker so live suites
+          // can import source modules that resolve siblings via a runtime
+          // `require("../module")` (e.g. inference/ollama-runtime-context.ts).
+          // Use setupFiles rather than NODE_OPTIONS so the hook stays in-process
+          // and never leaks `--require` into the real CLI subprocesses under
+          // test. Mirrors the `cli` project.
+          setupFiles: ["test/helpers/onboard-script-mocks.cjs"],
           testTimeout: testTimeout(LIVE_E2E_PROJECT_TIMEOUT_MS),
           // Live targets mutate host, Docker, gateway, and sandbox state. A
           // whole-test retry reuses that state and can hide the first failure