diff --git a/tests/e2e-cucumber/expectations.toml b/tests/e2e-cucumber/expectations.toml
index dde70512..0316bf62 100644
--- a/tests/e2e-cucumber/expectations.toml
+++ b/tests/e2e-cucumber/expectations.toml
@@ -10,7 +10,9 @@
 #   1. A scenario whose @requires-gpu / @requires-engine can't be satisfied on
 #      this host is SKIPPED (not-applicable) — never listed here.
 #   2. Otherwise, the FIRST matching condition below marks it expected-fail
-#      (xfail); a scenario that then PASSES is an XPASS (stale entry — remove it).
+#      (xfail); a scenario that then PASSES is an XPASS — a stale entry to be
+#      removed, UNLESS the row is `flaky = true`, which tolerates either result
+#      while the bug stays open (see the grammar below).
 #   3. No matching condition → expected-pass.
 #
 # Condition predicate grammar (deliberately tiny):
@@ -20,6 +22,8 @@
 #   bug = "EAI-NNNN"                  ticket reference (shown in the report)
 #   reason = "..."                    human explanation (shown in the report)
 #   serve_timeout_secs = 90          optional: shorter serve wait for a known bug
+#   flaky = true                     optional: bug reproduces only intermittently,
+#                                    so an XPASS is tolerated (reported, not stale)
 #
 # An empty `when = {}` means "always xfail" (platform-independent bug).
 
@@ -45,6 +49,16 @@ reason = "vLLM service reports ready (/v1/models 200) but POST /v1/chat/completi
 serve_timeout_secs = 90
 flaky = true
 
+# EAI-7423 was INTERMITTENT on Strix Halo Ubuntu as of 2026-08-05, not fixed:
+# which of its scenarios pass varied run to run (run 30995473578: 2 XPASS; run
+# 30993213188: 3 XPASS; run 30939967076: all 6). A deterministic-XPASS row fails
+# that lane on unrelated PRs, so the EAI-7423 lemonade-linux rows below carry
+# `flaky = true` — either outcome is tolerated while the bug stays open. The flag
+# is per-row and is NOT derived from `bug`, so a new EAI-7423 row needs it added
+# explicitly. Revisit under EAI-7853 before deleting any of them: which lemonade
+# build a given run used cannot currently be read back from this lane, so an
+# XPASS streak here is not by itself evidence that the bug is fixed.
+#
 # On a lemonade-native LINUX host (Strix Halo Ubuntu) the lemonade managed serve
 # loads the model and reaches ready on :8001, then is shut down ~0.08s later, so
 # the endpoint never becomes reachable (EAI-7423, root-caused on the box
@@ -56,6 +70,7 @@ when = { effective_engine = "lemonade", os = "linux" }
 bug = "EAI-7423"
 reason = "Lemonade managed serve reaches ready then shuts down immediately, so the endpoint never becomes reachable."
 serve_timeout_secs = 90
+flaky = true
 
 # On a lemonade-native LINUX host the served model's endpoint drops before
 # inference: the lemonade managed serve reaches ready then is shut down ~0.08s
@@ -66,6 +81,7 @@ when = { effective_engine = "lemonade", os = "linux" }
 bug = "EAI-7423"
 reason = "Lemonade managed serve reaches ready then shuts down immediately, so inference never succeeds."
 serve_timeout_secs = 90
+flaky = true
 
 # EAI-7333 (CLI healthcheck reports ready off /v1/models before inference works)
 # on the vLLM path. This scenario is FLAKY, not fixed: it XPASS'd on MI300X once
@@ -89,6 +105,7 @@ when = { effective_engine = "lemonade", os = "linux" }
 bug = "EAI-7423"
 reason = "Lemonade managed serve reaches ready then shuts down immediately, so the ready-to-serve contract fails."
 serve_timeout_secs = 90
+flaky = true
 
 # --- EAI-7423: on a lemonade LINUX host the managed serve reaches ready then is
 # shut down immediately, so inference never happens (root-caused on the box
@@ -100,6 +117,7 @@ when = { effective_engine = "lemonade", os = "linux" }
 bug = "EAI-7423"
 reason = "Lemonade managed serve reaches ready then shuts down immediately, so inference never succeeds."
 serve_timeout_secs = 90
+flaky = true
 
 # --- EAI-7223: chat requests with tool definitions rejected (vLLM path) ---
 [["chat-tool-definitions-accepted"]]
@@ -119,6 +137,7 @@ when = { effective_engine = "lemonade", os = "linux", therock_family = "gfx*" }
 bug = "EAI-7423"
 reason = "Lemonade managed serve reaches ready then shuts down immediately, so the chat request never reaches a live model."
 serve_timeout_secs = 90
+flaky = true
 
 # EAI-7221 (TUI chat sends hardcoded 'local-model' instead of querying /v1/models)
 # was previously mapped here as a vLLM-path xfail, but that bug is TUI-only and this
@@ -135,6 +154,7 @@ when = { effective_engine = "lemonade", os = "linux", therock_family = "gfx*" }
 bug = "EAI-7423"
 reason = "Lemonade managed serve reaches ready then shuts down immediately, so the chat request never reaches a live model."
 serve_timeout_secs = 90
+flaky = true
 
 # --- EAI-7383: `rocm help` lists subcommands in declaration order, not
 # alphabetically. Platform-independent (pure CLI help output). ---