From 00996ed71fbd3db3b9ef0b29dd5c1cebb442a49e Mon Sep 17 00:00:00 2001
From: Dustin Metzgar <dustin.metzgar@uipath.com>
Date: Wed, 23 Sep 2026 15:21:28 -0700
Subject: [PATCH 1/2] fix(tests): non-gating command telemetry must not move
 the score either
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

`pass_threshold: 0` was read as "this criterion is advisory", and
`test_v1_only_authoring_commands_match_the_temporary_allowlist` enforces only
that. It is half the idiom. coder_eval's own field documentation spells out the
other half:

    weight=0 excludes from the score but NOT from the pass/fail gate ...
    To make a criterion truly non-gating, also set pass_threshold=0.

So a criterion with `pass_threshold: 0` and `weight: 1.5` never fails a task and
always moves its score — it sits in both halves of the weighted mean.

That is not neutral between arms. A `command_executed` grades the SHELL COMMAND
an arm ran, and the two arms run different commands by construction. In the
2026-09-23 same-ground run, 15 such criteria across 10 flow tasks scored 1.0 for
v1 and 0.0 for v2 — `flow node add`, `flow registry get`, `agent init
--conversational`, `is connections list`, `is resources run list` — and dragged
tasks that passed every graded check down with them.
`datafabric_integration_create_get` returned SUCCESS at 0.55: three advisories
at weights 1.5/2.0/1.5 against two graded criteria at 3.0.

The outcome each of those stands in for is already graded by a `run_command`
against the artifact, which is route-blind. `slack_channel_description_simulated`
says so in its own comment: the run_command "Supersedes a static connector-key
grep ... THIS is the authoritative proof". So the telemetry keeps reporting and
stops scoring.

61 criteria across 38 tasks get `weight: 0` — three of them had no `weight` line
at all, which defaults to 1.0. `single_node/outlook_trigger_inbox` already did
this and is the shape the rest now match.

Two exclusions, both deliberate:

- `ixp/routing.yaml` — its two criteria carry `stop_early`, where weight is
  load-bearing for the pass-stop floor rather than for the score. Its own
  sentinel comment says so. The new gate skips `stop_early` criteria.
- `smoke/registry_discovery.yaml` — allowlisted. It "deliberately produces no
  artifact", so all four of its criteria grep the shell; zeroing them leaves a
  total weight of 0 and `calculate_weighted_score` reports 0.0 for that. It needs
  an outcome-graded criterion over the agent's report, which is a task redesign,
  not a reweight.

Replaying the run's recorded criterion results against the new weights: the flow
suite's v2 mean score goes 0.919 -> 0.935 and v1 0.959 -> 0.960, narrowing the
arm delta from -0.041 to -0.025. Pass/fail counts do not move, because weight
never gated. Seven v2 tasks reach a clean 1.00.

Two tasks per arm go DOWN, which is the point working: passing telemetry was
padding a failing grade. `e2e_devcon_expense_approval` drops 0.37 -> 0.23 in v2
and now reads as the HITL outcome-port failure it is.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
---
 .../_shared/test_same_ground_corpus.py        | 60 +++++++++++++++++++
 .../integration_create_get.yaml               |  6 +-
 .../datafabric_connector/smoke_error.yaml     |  4 +-
 .../datafabric_connector/smoke_query.yaml     |  2 +-
 .../connector_features/enum.yaml              |  2 +-
 .../generic_dynamic_node.yaml                 |  2 +-
 .../jdbc_databricks_query.yaml                |  2 +-
 .../paginated_reference_lookup.yaml           |  4 +-
 .../connector_features/path_params.yaml       |  2 +-
 .../slack_http_fallback.yaml                  |  2 +-
 .../testmanager_attachments.yaml              |  2 +-
 .../testmanager_crud_grounded.yaml            |  2 +-
 .../testmanager_execution_results.yaml        |  2 +-
 .../testmanager_generic_records.yaml          |  2 +-
 .../testmanager_requirement_lifecycle.yaml    |  2 +-
 .../testmanager_testcase_lifecycle.yaml       |  2 +-
 .../testmanager_testset_lifecycle.yaml        |  2 +-
 .../webhook_waitfor_parallel.yaml             |  4 +-
 .../conversational_chat_loop.yaml             |  6 +-
 .../e2e/devcon_expense_approval.yaml          |  6 +-
 .../e2e/jira_lifecycle/jira_lifecycle.yaml    |  2 +-
 .../evaluate/evaluator_type_choice.yaml       |  2 +-
 .../evaluate/local_crud.yaml                  |  6 +-
 .../evaluate/no_auto_upload.yaml              |  2 +-
 .../evaluate/simulation/simulation_crud.yaml  |  6 +-
 .../customer_escalation_triage.yaml           |  2 +-
 .../slack_channel_description_simulated.yaml  |  2 +-
 .../ixp/e2e_02_project_selection.yaml         |  2 +-
 .../e2e_03_project_creation_handoff.yaml      |  4 +-
 .../ixp/e2e_04_build_mechanics.yaml           |  4 +-
 .../ixp/integration_handle_routing.yaml       |  2 +-
 .../ixp/routing_listing.yaml                  |  1 +
 .../ixp/scaffold_minimal.yaml                 |  4 +-
 .../ixp/scaffold_multinode.yaml               |  4 +-
 .../integration_native_read_create.yaml       |  2 +-
 .../smoke/init_maestro_automate.yaml          |  2 +-
 .../smoke/init_validate.yaml                  |  2 +-
 .../voice/voice_inbound_call.yaml             |  2 +-
 .../voice/voice_outbound_call.yaml            |  6 +-
 39 files changed, 117 insertions(+), 56 deletions(-)

diff --git a/tests/tasks/uipath-maestro-flow/_shared/test_same_ground_corpus.py b/tests/tasks/uipath-maestro-flow/_shared/test_same_ground_corpus.py
index 1b0ead6336..6917ca6168 100644
--- a/tests/tasks/uipath-maestro-flow/_shared/test_same_ground_corpus.py
+++ b/tests/tasks/uipath-maestro-flow/_shared/test_same_ground_corpus.py
@@ -42,6 +42,22 @@
 
 DEBUG_SOLUTION_ALLOWLIST = set()
 
+# Non-gating `command_executed` criteria must also be weightless — see
+# `test_non_gating_command_telemetry_is_weightless`.
+#
+# `smoke/registry_discovery` is the one task whose ENTIRE grade is command
+# telemetry: it "deliberately produces no artifact", so all four of its criteria
+# grep the shell. Zeroing them leaves a total weight of 0, and
+# `calculate_weighted_score` reports 0.0 for that — the task would score nothing
+# whatever the agent did. The real fix is an outcome-graded criterion over the
+# agent's REPORT (the prompt asks it to name the two node types and their
+# schemas), which is a task redesign rather than a reweight. Until then it keeps
+# its weights and stays arm-biased by construction: an SDK-loop agent can answer
+# this from the SDK's own `api-index.md` without touching `flow registry` at all.
+NON_GATING_WEIGHT_ALLOWLIST = {
+    "smoke/registry_discovery.yaml",
+}
+
 # The two billing lookups name only a "Data Service entity", which since #3041
 # denotes two node families. Their prompts pin the connector so the graded
 # structure is deterministic; the sibling dispute-resolution task pins it in the
@@ -184,6 +200,50 @@ def test_v1_only_authoring_commands_match_the_temporary_allowlist() -> None:
     assert offenders == V1_AUTHORING_ALLOWLIST
 
 
+def test_non_gating_command_telemetry_is_weightless() -> None:
+    """A `command_executed` that does not gate must not move the score either.
+
+    THE GAP THIS EXISTS FOR. `pass_threshold: 0` was read as "this criterion is
+    advisory", and the sibling test above enforces only that. It is half the
+    idiom. coder_eval's own field docs spell out the other half: "weight=0
+    excludes from the score but NOT from the pass/fail gate ... To make a
+    criterion truly non-gating, also set pass_threshold=0." A criterion with
+    `pass_threshold: 0` and `weight: 1.5` still lands in both halves of the
+    weighted mean, so failing it costs score without ever failing the task.
+
+    That is not neutral between arms, because a `command_executed` grades the
+    SHELL COMMAND an arm ran, and the two arms run different commands by
+    construction. In the 2026-09-23 same-ground run, 15 such criteria across
+    10 flow tasks scored 1.0 for v1 and 0.0 for v2 — `flow node add`,
+    `flow registry get`, `agent init --conversational`, `is connections list`
+    — dragging tasks that passed every graded check down to 0.55, 0.59, 0.71.
+    `datafabric_integration_create_get` returned SUCCESS at 0.55.
+
+    The outcome those criteria stand in for is graded by a `run_command`
+    against the artifact, which is route-blind. So the telemetry keeps
+    reporting and stops scoring.
+
+    Criteria carrying `stop_early` are exempt: there `weight` is load-bearing
+    for the pass-stop floor, not just for the score (see `ixp/routing.yaml`,
+    whose sentinel says so).
+    """
+    offenders = set()
+    for relative, _, text in _tagged_tasks():
+        for criterion_type, criterion in _criterion_blocks(text):
+            if criterion_type != "command_executed":
+                continue
+            threshold = re.search(r"(?m)^\s+pass_threshold:\s*([0-9.]+)", criterion)
+            if threshold is None or float(threshold.group(1)) > 0:
+                continue
+            if re.search(r"(?m)^\s+stop_early:", criterion):
+                continue
+            weight = re.search(r"(?m)^\s+weight:\s*([0-9.]+)", criterion)
+            # An absent `weight` defaults to 1.0, so silence is not compliance.
+            if weight is None or float(weight.group(1)) != 0:
+                offenders.add(relative)
+    assert offenders == NON_GATING_WEIGHT_ALLOWLIST
+
+
 def test_gating_skill_telemetry_matches_the_temporary_allowlist() -> None:
     offenders = set()
     for relative, _, text in _tagged_tasks():
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/datafabric_connector/integration_create_get.yaml b/tests/tasks/uipath-maestro-flow/connector_features/datafabric_connector/integration_create_get.yaml
index f2b74107e4..dee42a2f41 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/datafabric_connector/integration_create_get.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/datafabric_connector/integration_create_get.yaml
@@ -76,7 +76,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+node\s+add\s+\S+\s+"?uipath\.connector\.uipath-uipath-dataservice\.create-entity-record'
     min_count: 1
-    weight: 1.5
+    weight: 0
     pass_threshold: 0.0
 
   - type: command_executed
@@ -84,7 +84,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+node\s+add\s+\S+\s+"?uipath\.connector\.uipath-uipath-dataservice\.get-entity-record-by-id'
     min_count: 1
-    weight: 2.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: command_executed
@@ -92,7 +92,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+node\s+add\s+\S+\s+"?uipath\.connector\.uipath-uipath-dataservice\.delete-entity-record'
     min_count: 1
-    weight: 1.5
+    weight: 0
     pass_threshold: 0.0
 
   - type: command_executed
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/datafabric_connector/smoke_error.yaml b/tests/tasks/uipath-maestro-flow/connector_features/datafabric_connector/smoke_error.yaml
index 0798c3f199..7c368a1307 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/datafabric_connector/smoke_error.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/datafabric_connector/smoke_error.yaml
@@ -61,14 +61,14 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+node\s+add\s+\S+\s+"?uipath\.connector\.uipath-uipath-dataservice\.create-entity-record'
     min_count: 1
-    weight: 1.5
+    weight: 0
     pass_threshold: 0.0
   - type: command_executed
     description: "Advisory: live-v1 agent added the follow-up Query nodes"
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+node\s+add\s+\S+\s+"?uipath\.connector\.uipath-uipath-dataservice\.query-entity-records'
     min_count: 1
-    weight: 2.0
+    weight: 0
     pass_threshold: 0.0
   - type: command_executed
     description: "Flow validated"
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/datafabric_connector/smoke_query.yaml b/tests/tasks/uipath-maestro-flow/connector_features/datafabric_connector/smoke_query.yaml
index ba82b3de0b..be27d1a5e7 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/datafabric_connector/smoke_query.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/datafabric_connector/smoke_query.yaml
@@ -78,7 +78,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+node\s+add\s+\S+\s+"?uipath\.connector\.uipath-uipath-dataservice\.query-entity-records'
     min_count: 1
-    weight: 2.0
+    weight: 0
     pass_threshold: 0.0
   - type: command_executed
     description: "Flow validated"
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/enum.yaml b/tests/tasks/uipath-maestro-flow/connector_features/enum.yaml
index 29d823f98b..551e703dee 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/enum.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/enum.yaml
@@ -44,7 +44,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: "(uip|\\$UIP)\\s+is\\s+connections\\s+list"
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/generic_dynamic_node/generic_dynamic_node.yaml b/tests/tasks/uipath-maestro-flow/connector_features/generic_dynamic_node/generic_dynamic_node.yaml
index 4b2b88d77e..7df3ce3aa4 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/generic_dynamic_node/generic_dynamic_node.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/generic_dynamic_node/generic_dynamic_node.yaml
@@ -79,7 +79,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP|\$\{?UIP\}?)\s+(maestro\s+)?flow\s+debug'
     min_count: 1
-    weight: 1.5
+    weight: 0
     pass_threshold: 0.0
 
   # ── Execution: connector calls ServiceNow and surfaces an array output ──
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/jdbc_databricks_query/jdbc_databricks_query.yaml b/tests/tasks/uipath-maestro-flow/connector_features/jdbc_databricks_query/jdbc_databricks_query.yaml
index d53cf81b50..2263dde29c 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/jdbc_databricks_query/jdbc_databricks_query.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/jdbc_databricks_query/jdbc_databricks_query.yaml
@@ -70,7 +70,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+(maestro\s+)?flow\s+debug'
     min_count: 1
-    weight: 1.5
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/paginated_reference_lookup.yaml b/tests/tasks/uipath-maestro-flow/connector_features/paginated_reference_lookup.yaml
index 4d944a41dc..2f181a0cdd 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/paginated_reference_lookup.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/paginated_reference_lookup.yaml
@@ -65,7 +65,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: 'uip\s+is\s+resources\s+run\s+list\s+\\?"?uipath-salesforce-slack\\?"?\s+\\?"?(curated_channels|conversations)'
     min_count: 2
-    weight: 3.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: command_executed
@@ -74,7 +74,7 @@ success_criteria:
     command_pattern: '(uip\s+is\s+resources\s+run\s+list\s+\\?"?uipath-salesforce-slack\\?"?\s+\\?"?(curated_channels|conversations)[^\n]*nextPage=|registry\s+prepare\s+\\?"?uipath-salesforce-slack\\?"?[^\n]*--resolve\s+\\?"?channel:)'
     min_count: 1
     require_success: true
-    weight: 3.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: command_executed
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/path_params.yaml b/tests/tasks/uipath-maestro-flow/connector_features/path_params.yaml
index 10cae23547..450e5bde4f 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/path_params.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/path_params.yaml
@@ -46,7 +46,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: "(uip|\\$UIP)\\s+is\\s+connections\\s+list"
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/slack-http-fallback/slack_http_fallback.yaml b/tests/tasks/uipath-maestro-flow/connector_features/slack-http-fallback/slack_http_fallback.yaml
index 90562fe98c..334b17e4ba 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/slack-http-fallback/slack_http_fallback.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/slack-http-fallback/slack_http_fallback.yaml
@@ -56,7 +56,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: "(uip|\\$UIP)\\s+(maestro\\s+)?flow\\s+debug"
     min_count: 1
-    weight: 1.5
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/testmanager_attachments/testmanager_attachments.yaml b/tests/tasks/uipath-maestro-flow/connector_features/testmanager_attachments/testmanager_attachments.yaml
index 8b7d1abeaf..8c83464199 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/testmanager_attachments/testmanager_attachments.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/testmanager_attachments/testmanager_attachments.yaml
@@ -52,7 +52,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+registry\s+pull'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/testmanager_crud_grounded/testmanager_crud_grounded.yaml b/tests/tasks/uipath-maestro-flow/connector_features/testmanager_crud_grounded/testmanager_crud_grounded.yaml
index 8ee194ad59..0afe281322 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/testmanager_crud_grounded/testmanager_crud_grounded.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/testmanager_crud_grounded/testmanager_crud_grounded.yaml
@@ -52,7 +52,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+registry\s+pull'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/testmanager_execution_results/testmanager_execution_results.yaml b/tests/tasks/uipath-maestro-flow/connector_features/testmanager_execution_results/testmanager_execution_results.yaml
index 6a74bf9d29..22cfe2a506 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/testmanager_execution_results/testmanager_execution_results.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/testmanager_execution_results/testmanager_execution_results.yaml
@@ -57,7 +57,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+registry\s+pull'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/testmanager_generic_records/testmanager_generic_records.yaml b/tests/tasks/uipath-maestro-flow/connector_features/testmanager_generic_records/testmanager_generic_records.yaml
index 5a76c3e868..ec630950b8 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/testmanager_generic_records/testmanager_generic_records.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/testmanager_generic_records/testmanager_generic_records.yaml
@@ -53,7 +53,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+registry\s+pull'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/testmanager_requirement_lifecycle/testmanager_requirement_lifecycle.yaml b/tests/tasks/uipath-maestro-flow/connector_features/testmanager_requirement_lifecycle/testmanager_requirement_lifecycle.yaml
index 4067533d40..e468906892 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/testmanager_requirement_lifecycle/testmanager_requirement_lifecycle.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/testmanager_requirement_lifecycle/testmanager_requirement_lifecycle.yaml
@@ -54,7 +54,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+registry\s+pull'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/testmanager_testcase_lifecycle/testmanager_testcase_lifecycle.yaml b/tests/tasks/uipath-maestro-flow/connector_features/testmanager_testcase_lifecycle/testmanager_testcase_lifecycle.yaml
index 95ae87c3fe..0f7970df24 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/testmanager_testcase_lifecycle/testmanager_testcase_lifecycle.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/testmanager_testcase_lifecycle/testmanager_testcase_lifecycle.yaml
@@ -53,7 +53,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+registry\s+pull'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/testmanager_testset_lifecycle/testmanager_testset_lifecycle.yaml b/tests/tasks/uipath-maestro-flow/connector_features/testmanager_testset_lifecycle/testmanager_testset_lifecycle.yaml
index 5e462aec9e..e22e852f8d 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/testmanager_testset_lifecycle/testmanager_testset_lifecycle.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/testmanager_testset_lifecycle/testmanager_testset_lifecycle.yaml
@@ -55,7 +55,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+registry\s+pull'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/connector_trigger/webhook_waitfor_parallel.yaml b/tests/tasks/uipath-maestro-flow/connector_trigger/webhook_waitfor_parallel.yaml
index 0821a6a271..0dc9a9d4fd 100644
--- a/tests/tasks/uipath-maestro-flow/connector_trigger/webhook_waitfor_parallel.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_trigger/webhook_waitfor_parallel.yaml
@@ -69,7 +69,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+solution\s+(new|init)'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
@@ -85,7 +85,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+is\s+webhooks\s+config'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/conversational/conversational_chat_loop.yaml b/tests/tasks/uipath-maestro-flow/conversational/conversational_chat_loop.yaml
index b9c9ed4516..08dcbdf3f1 100644
--- a/tests/tasks/uipath-maestro-flow/conversational/conversational_chat_loop.yaml
+++ b/tests/tasks/uipath-maestro-flow/conversational/conversational_chat_loop.yaml
@@ -121,7 +121,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+agent\s+init\s+.*--conversational'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0
 
   - type: command_executed
@@ -129,7 +129,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+agent\s+refresh\s+.*--inline-in-flow'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0
 
   - type: command_executed
@@ -137,5 +137,5 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+.*--output\s+json'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0
diff --git a/tests/tasks/uipath-maestro-flow/e2e/devcon_expense_approval.yaml b/tests/tasks/uipath-maestro-flow/e2e/devcon_expense_approval.yaml
index 898f3639b6..b12fd061fa 100644
--- a/tests/tasks/uipath-maestro-flow/e2e/devcon_expense_approval.yaml
+++ b/tests/tasks/uipath-maestro-flow/e2e/devcon_expense_approval.yaml
@@ -56,7 +56,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP|\$\{?UIP\}?)\s+solution\s+(new|init)'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0  # advisory: init auto-scaffolds the parent solution (CLI #2470)
 
   - type: command_executed
@@ -64,7 +64,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP|\$\{?UIP\}?)\s+(maestro\s+)?flow\s+init'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
@@ -88,7 +88,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP|\$\{?UIP\}?)\s+(maestro\s+)?flow\s+validate'
     min_count: 1
-    weight: 2.5
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/e2e/jira_lifecycle/jira_lifecycle.yaml b/tests/tasks/uipath-maestro-flow/e2e/jira_lifecycle/jira_lifecycle.yaml
index a58480e5c9..7c4a858286 100644
--- a/tests/tasks/uipath-maestro-flow/e2e/jira_lifecycle/jira_lifecycle.yaml
+++ b/tests/tasks/uipath-maestro-flow/e2e/jira_lifecycle/jira_lifecycle.yaml
@@ -89,7 +89,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP|\$\{?UIP\}?)\s+(maestro\s+)?flow\s+debug'
     min_count: 1
-    weight: 1.5
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/evaluate/evaluator_type_choice.yaml b/tests/tasks/uipath-maestro-flow/evaluate/evaluator_type_choice.yaml
index f536ae4b9d..23ee88c538 100644
--- a/tests/tasks/uipath-maestro-flow/evaluate/evaluator_type_choice.yaml
+++ b/tests/tasks/uipath-maestro-flow/evaluate/evaluator_type_choice.yaml
@@ -67,7 +67,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+solution\s+(new|init)'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0  # advisory: init auto-scaffolds the parent solution (CLI #2470)
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/evaluate/local_crud.yaml b/tests/tasks/uipath-maestro-flow/evaluate/local_crud.yaml
index ca3ea68415..680d107b75 100644
--- a/tests/tasks/uipath-maestro-flow/evaluate/local_crud.yaml
+++ b/tests/tasks/uipath-maestro-flow/evaluate/local_crud.yaml
@@ -58,7 +58,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+solution\s+(new|init)'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0  # advisory: init auto-scaffolds the parent solution (CLI #2470)
 
   - type: command_executed
@@ -66,7 +66,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+(maestro\s+)?flow\s+init'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
@@ -106,7 +106,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+eval\s+(evaluator\s+list|set\s+list)'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/evaluate/no_auto_upload.yaml b/tests/tasks/uipath-maestro-flow/evaluate/no_auto_upload.yaml
index aadaa966eb..77645c39f4 100644
--- a/tests/tasks/uipath-maestro-flow/evaluate/no_auto_upload.yaml
+++ b/tests/tasks/uipath-maestro-flow/evaluate/no_auto_upload.yaml
@@ -65,7 +65,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+solution\s+(new|init)'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0  # advisory: init auto-scaffolds the parent solution (CLI #2470)
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/evaluate/simulation/simulation_crud.yaml b/tests/tasks/uipath-maestro-flow/evaluate/simulation/simulation_crud.yaml
index 1468ac2d3e..4f23b449b5 100644
--- a/tests/tasks/uipath-maestro-flow/evaluate/simulation/simulation_crud.yaml
+++ b/tests/tasks/uipath-maestro-flow/evaluate/simulation/simulation_crud.yaml
@@ -67,7 +67,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+solution\s+(new|init)'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
@@ -107,7 +107,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+eval\s+simulation\s+add\s+(?=.*--strategy\s+Static\b)(?=.*--mock-value\b)'
     min_count: 1
-    weight: 3.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
@@ -123,7 +123,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+eval\s+simulation\s+list'
     min_count: 1
-    weight: 1.5
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/interactive/customer_escalation_triage/customer_escalation_triage.yaml b/tests/tasks/uipath-maestro-flow/interactive/customer_escalation_triage/customer_escalation_triage.yaml
index db1ada8b25..551705ac81 100644
--- a/tests/tasks/uipath-maestro-flow/interactive/customer_escalation_triage/customer_escalation_triage.yaml
+++ b/tests/tasks/uipath-maestro-flow/interactive/customer_escalation_triage/customer_escalation_triage.yaml
@@ -92,7 +92,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP|\$\{?UIP\}?)\s+(maestro\s+)?flow\s+validate'
     min_count: 1
-    weight: 1.5
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/interactive/slack_channel_description_simulated/slack_channel_description_simulated.yaml b/tests/tasks/uipath-maestro-flow/interactive/slack_channel_description_simulated/slack_channel_description_simulated.yaml
index bfffd79a23..b30e7c331d 100644
--- a/tests/tasks/uipath-maestro-flow/interactive/slack_channel_description_simulated/slack_channel_description_simulated.yaml
+++ b/tests/tasks/uipath-maestro-flow/interactive/slack_channel_description_simulated/slack_channel_description_simulated.yaml
@@ -88,7 +88,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP|\$\{?UIP\}?)\s+(maestro\s+)?flow\s+debug'
     min_count: 1
-    weight: 1.5
+    weight: 0
     pass_threshold: 0.0
 
   # Runtime check (mirrors the retired non-simulated original, check_channel_description.py):
diff --git a/tests/tasks/uipath-maestro-flow/ixp/e2e_02_project_selection.yaml b/tests/tasks/uipath-maestro-flow/ixp/e2e_02_project_selection.yaml
index 6e86658646..5e1b045c25 100644
--- a/tests/tasks/uipath-maestro-flow/ixp/e2e_02_project_selection.yaml
+++ b/tests/tasks/uipath-maestro-flow/ixp/e2e_02_project_selection.yaml
@@ -77,7 +77,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: 'uip\s+maestro\s+flow\s+registry\s+(search|list|get)\b[^\n]*ixp'
     min_count: 1
-    weight: 2.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/ixp/e2e_03_project_creation_handoff/e2e_03_project_creation_handoff.yaml b/tests/tasks/uipath-maestro-flow/ixp/e2e_03_project_creation_handoff/e2e_03_project_creation_handoff.yaml
index dd23318249..4cf0a3edda 100644
--- a/tests/tasks/uipath-maestro-flow/ixp/e2e_03_project_creation_handoff/e2e_03_project_creation_handoff.yaml
+++ b/tests/tasks/uipath-maestro-flow/ixp/e2e_03_project_creation_handoff/e2e_03_project_creation_handoff.yaml
@@ -145,7 +145,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(?i)(uip|\$UIP)\s+maestro\s+flow\s+registry\s+search'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0
 
   # Advisory until a real run proves it latches. impl.md requires the
@@ -157,5 +157,5 @@ success_criteria:
     description: "Advisory: handoff payload names the documents folder"
     command_pattern: 'uipath-ixp[\s\S]{0,600}documents'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0
diff --git a/tests/tasks/uipath-maestro-flow/ixp/e2e_04_build_mechanics.yaml b/tests/tasks/uipath-maestro-flow/ixp/e2e_04_build_mechanics.yaml
index 1ad52ee475..f5373060bb 100644
--- a/tests/tasks/uipath-maestro-flow/ixp/e2e_04_build_mechanics.yaml
+++ b/tests/tasks/uipath-maestro-flow/ixp/e2e_04_build_mechanics.yaml
@@ -123,7 +123,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(?i)(uip|\$UIP)\s+ixp\s+projects\s+create'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0
 
   - type: command_executed
@@ -131,7 +131,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(?i)(uip|\$UIP)\s+ixp\s+deployments\s+create\b[^\n]*--folder-key'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0
 
   # Primary gate: project created + folder-deployed this run, then the .flow
diff --git a/tests/tasks/uipath-maestro-flow/ixp/integration_handle_routing.yaml b/tests/tasks/uipath-maestro-flow/ixp/integration_handle_routing.yaml
index 6ff7381e77..807c7f4a28 100644
--- a/tests/tasks/uipath-maestro-flow/ixp/integration_handle_routing.yaml
+++ b/tests/tasks/uipath-maestro-flow/ixp/integration_handle_routing.yaml
@@ -69,7 +69,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: 'uip\s+maestro\s+flow\s+registry\s+(search|list|get)\b[^\n]*ixp'
     min_count: 1
-    weight: 2.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/ixp/routing_listing.yaml b/tests/tasks/uipath-maestro-flow/ixp/routing_listing.yaml
index 4019ecf40e..7ec546268d 100644
--- a/tests/tasks/uipath-maestro-flow/ixp/routing_listing.yaml
+++ b/tests/tasks/uipath-maestro-flow/ixp/routing_listing.yaml
@@ -65,6 +65,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(?i)uip\s+maestro\s+flow\s+registry\s+(list\b|(search|list)\b[^\n]*\b(ixp|document|extract))'
     min_count: 1
+    weight: 0
     pass_threshold: 0.0
 
   # A listing/Q&A task should not create a Flow artifact.
diff --git a/tests/tasks/uipath-maestro-flow/ixp/scaffold_minimal.yaml b/tests/tasks/uipath-maestro-flow/ixp/scaffold_minimal.yaml
index 694d23148b..0d30eff452 100644
--- a/tests/tasks/uipath-maestro-flow/ixp/scaffold_minimal.yaml
+++ b/tests/tasks/uipath-maestro-flow/ixp/scaffold_minimal.yaml
@@ -58,7 +58,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(?i)uip\s+maestro\s+flow\s+registry\s+(list\b|(search|list|get)\b[^\n]*\b(ixp|document|extract))'
     min_count: 1
-    weight: 2.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: command_executed
@@ -66,7 +66,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: 'uip\s+(maestro\s+)?flow\s+validate'
     min_count: 1
-    weight: 1.5
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/ixp/scaffold_multinode.yaml b/tests/tasks/uipath-maestro-flow/ixp/scaffold_multinode.yaml
index 5213c520df..671b639c78 100644
--- a/tests/tasks/uipath-maestro-flow/ixp/scaffold_multinode.yaml
+++ b/tests/tasks/uipath-maestro-flow/ixp/scaffold_multinode.yaml
@@ -62,7 +62,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(?i)uip\s+maestro\s+flow\s+registry\s+(list\b|(search|list|get)\b[^\n]*\b(ixp|document|extract))'
     min_count: 1
-    weight: 2.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: command_executed
@@ -70,7 +70,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: 'uip\s+(maestro\s+)?flow\s+validate'
     min_count: 1
-    weight: 1.5
+    weight: 0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/node_features/datafabric_native/integration_native_read_create.yaml b/tests/tasks/uipath-maestro-flow/node_features/datafabric_native/integration_native_read_create.yaml
index ecda35c398..66d5f1ebe3 100644
--- a/tests/tasks/uipath-maestro-flow/node_features/datafabric_native/integration_native_read_create.yaml
+++ b/tests/tasks/uipath-maestro-flow/node_features/datafabric_native/integration_native_read_create.yaml
@@ -69,7 +69,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+registry\s+get\s+"?core\.datafabric\.'
     min_count: 1
-    weight: 2.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: command_executed
diff --git a/tests/tasks/uipath-maestro-flow/smoke/init_maestro_automate.yaml b/tests/tasks/uipath-maestro-flow/smoke/init_maestro_automate.yaml
index 0660dec09d..8aa9a5073e 100644
--- a/tests/tasks/uipath-maestro-flow/smoke/init_maestro_automate.yaml
+++ b/tests/tasks/uipath-maestro-flow/smoke/init_maestro_automate.yaml
@@ -35,7 +35,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+(maestro\s+)?flow\s+init\b(?=(?:(?!\n|&&|\|\||;|\||\s(?:uip|\$UIP)\s).)*--automate)'
     min_count: 1
-    weight: 1.5
+    weight: 0
     pass_threshold: 0.0
 
   # Scoped to the project the task asked for. A sandbox-wide search passes
diff --git a/tests/tasks/uipath-maestro-flow/smoke/init_validate.yaml b/tests/tasks/uipath-maestro-flow/smoke/init_validate.yaml
index fa1f91fa18..93efcc725e 100644
--- a/tests/tasks/uipath-maestro-flow/smoke/init_validate.yaml
+++ b/tests/tasks/uipath-maestro-flow/smoke/init_validate.yaml
@@ -34,7 +34,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+solution\s+(init|new)'
     min_count: 1
-    weight: 1.5
+    weight: 0
     pass_threshold: 0.0  # advisory: init auto-scaffolds the parent solution (CLI #2470)
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/voice/voice_inbound_call.yaml b/tests/tasks/uipath-maestro-flow/voice/voice_inbound_call.yaml
index e97a1311d8..b3eba1dd4e 100644
--- a/tests/tasks/uipath-maestro-flow/voice/voice_inbound_call.yaml
+++ b/tests/tasks/uipath-maestro-flow/voice/voice_inbound_call.yaml
@@ -151,5 +151,5 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+agent\s+init\s+.*--conversational'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0
diff --git a/tests/tasks/uipath-maestro-flow/voice/voice_outbound_call.yaml b/tests/tasks/uipath-maestro-flow/voice/voice_outbound_call.yaml
index 51343b1070..dea9283907 100644
--- a/tests/tasks/uipath-maestro-flow/voice/voice_outbound_call.yaml
+++ b/tests/tasks/uipath-maestro-flow/voice/voice_outbound_call.yaml
@@ -129,7 +129,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+agent\s+init\s+.*--conversational'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0
 
   - type: command_executed
@@ -137,7 +137,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+agent\s+refresh\s+.*--inline-in-flow'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0
 
   - type: command_executed
@@ -145,5 +145,5 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+.*--output\s+json'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0

From 846df497741c6b0f40265a74095f5b6a33b54c1b Mon Sep 17 00:00:00 2001
From: Dustin Metzgar <dustin.metzgar@uipath.com>
Date: Wed, 23 Sep 2026 16:44:40 -0700
Subject: [PATCH 2/2] fix(tests): key the reweight on the COMMAND, not the
 criterion type
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Review caught this, and the review was right. The first pass keyed on criterion
TYPE — every non-gating `command_executed` — on the reasoning that such a
criterion grades the shell command an arm ran, and the two arms run different
commands. That reasoning does not survive the data.

Of the 54 criteria it swept that the 2026-09-23 run actually observed, 31 were
BOTH PASS: `solution init` (5), `flow validate` (4), `flow debug` (3),
`flow eval ...` (3), `flow init` (2), `--output json`, `is webhooks config`,
`ixp projects create`. Those are lifecycle commands both routes run on the same
artifact, and both routes do run them. Zeroing them removed real, satisfiable
signal and bought no neutrality — and it moved v1's own mean, which a neutrality
correction has no business doing.

So the rule is now about the COMMAND: a criterion is in scope only when its
command has no counterpart in the other route. Four families, each with its
reason recorded next to it:

  - `flow node add|configure|remove|update`, `flow edge ...` — v1 mutates the
    graph a node at a time; the SDK loop writes `.flow.ts`.
  - `agent init|refresh --inline-in-flow|--conversational` — v1 scaffolds the
    sidecar with the CLI; `conversationalAgent()` emits `agent.json` itself.
  - `is connections list`, `is resources run list`, `is triggers objects|describe`
    — `registry prepare` picks the connection, pages the collection and writes
    bindings in one call.
  - `flow registry pull` — the manifest refresh the SDK loop has no need of.

`flow registry get|search|list` is deliberately NOT a family: the SDK arm passes
those in the IxP tasks, so the node registry is not a v1-only surface. Only the
refresh step is.

23 criteria across 17 files, down from 61 across 38.

Replayed against the run's recorded results: v2's mean goes 0.919 -> 0.933 and
the arm delta -0.041 -> -0.027, within a thousandth of the broad sweep's 0.935.
v1's mean does not move at all and no task in either arm goes down — where the
first pass dropped `e2e_devcon_expense_approval` 0.37 -> 0.23 and
`interactive_customer_escalation_triage` 0.45 -> 0.38 by zeroing validate and
solution-init telemetry those tasks were passing.

`smoke/registry_discovery` no longer needs an allowlist: only its `registry
pull` criterion is in scope, leaving 4.5 of weight behind, so the zero-total
score the broad sweep would have produced cannot arise.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
---
 .../_shared/test_same_ground_corpus.py        | 105 +++++++++++-------
 .../generic_dynamic_node.yaml                 |   2 +-
 .../jdbc_databricks_query.yaml                |   2 +-
 .../slack_http_fallback.yaml                  |   2 +-
 .../webhook_waitfor_parallel.yaml             |   4 +-
 .../conversational_chat_loop.yaml             |   2 +-
 .../e2e/devcon_expense_approval.yaml          |   6 +-
 .../e2e/jira_lifecycle/jira_lifecycle.yaml    |   2 +-
 .../evaluate/evaluator_type_choice.yaml       |   2 +-
 .../evaluate/local_crud.yaml                  |   6 +-
 .../evaluate/no_auto_upload.yaml              |   2 +-
 .../evaluate/simulation/simulation_crud.yaml  |   6 +-
 .../customer_escalation_triage.yaml           |   2 +-
 .../slack_channel_description_simulated.yaml  |   2 +-
 .../ixp/e2e_02_project_selection.yaml         |   2 +-
 .../e2e_03_project_creation_handoff.yaml      |   4 +-
 .../ixp/e2e_04_build_mechanics.yaml           |   4 +-
 .../ixp/integration_handle_routing.yaml       |   2 +-
 .../ixp/routing_listing.yaml                  |   1 -
 .../ixp/scaffold_minimal.yaml                 |   4 +-
 .../ixp/scaffold_multinode.yaml               |   4 +-
 .../integration_native_read_create.yaml       |   2 +-
 .../smoke/init_maestro_automate.yaml          |   2 +-
 .../smoke/init_validate.yaml                  |   2 +-
 .../smoke/registry_discovery.yaml             |   2 +-
 .../voice/voice_outbound_call.yaml            |   2 +-
 26 files changed, 101 insertions(+), 75 deletions(-)

diff --git a/tests/tasks/uipath-maestro-flow/_shared/test_same_ground_corpus.py b/tests/tasks/uipath-maestro-flow/_shared/test_same_ground_corpus.py
index 6917ca6168..993be8f1b4 100644
--- a/tests/tasks/uipath-maestro-flow/_shared/test_same_ground_corpus.py
+++ b/tests/tasks/uipath-maestro-flow/_shared/test_same_ground_corpus.py
@@ -42,21 +42,22 @@
 
 DEBUG_SOLUTION_ALLOWLIST = set()
 
-# Non-gating `command_executed` criteria must also be weightless — see
-# `test_non_gating_command_telemetry_is_weightless`.
-#
-# `smoke/registry_discovery` is the one task whose ENTIRE grade is command
-# telemetry: it "deliberately produces no artifact", so all four of its criteria
-# grep the shell. Zeroing them leaves a total weight of 0, and
-# `calculate_weighted_score` reports 0.0 for that — the task would score nothing
-# whatever the agent did. The real fix is an outcome-graded criterion over the
-# agent's REPORT (the prompt asks it to name the two node types and their
-# schemas), which is a task redesign rather than a reweight. Until then it keeps
-# its weights and stays arm-biased by construction: an SDK-loop agent can answer
-# this from the SDK's own `api-index.md` without touching `flow registry` at all.
-NON_GATING_WEIGHT_ALLOWLIST = {
-    "smoke/registry_discovery.yaml",
-}
+# Commands whose JOB the other route does differently or internally, so a
+# `command_executed` on one of them measures which route ran rather than what
+# the run produced. See `test_route_specific_command_telemetry_is_weightless`
+# for the evidence behind each family, and for what is deliberately NOT here.
+ROUTE_SPECIFIC_COMMANDS = (
+    # v1 mutates the graph a node at a time; the SDK loop writes `.flow.ts`.
+    r"flow node \(?(add|configure|remove|update)|flow edge ",
+    # v1 scaffolds the inline agent's sidecar with the CLI; the SDK's
+    # `conversationalAgent()` / `agent()` emit `agent.json` themselves.
+    r"agent (init|refresh) (?=.*(inline-in-flow|--conversational))",
+    # v1 walks the tenant by hand; `registry prepare` picks the connection,
+    # pages the collection and writes `bindings.json` in one call.
+    r"is connections list|is resources run list|is triggers \(?(objects|describe)",
+    # v1 refreshes the node manifest before searching it.
+    r"flow registry \(?pull",
+)
 
 # The two billing lookups name only a "Data Service entity", which since #3041
 # denotes two node families. Their prompts pin the connector so the graded
@@ -200,32 +201,55 @@ def test_v1_only_authoring_commands_match_the_temporary_allowlist() -> None:
     assert offenders == V1_AUTHORING_ALLOWLIST
 
 
-def test_non_gating_command_telemetry_is_weightless() -> None:
-    """A `command_executed` that does not gate must not move the score either.
+def _command_pattern(criterion: str) -> str:
+    """A `command_pattern` with its regex escaping flattened to plain words.
+
+    The corpus spells the same command several ways — `\\s+` in a single-quoted
+    scalar, `\\\\s+` in a double-quoted one — so matching families against the
+    raw text would miss half of them.
+    """
+    match = re.search(r"(?m)^\s+command_pattern:\s*(.*)$", criterion)
+    if match is None:
+        return ""
+    pattern = match.group(1).replace("\\\\", "\\")
+    pattern = re.sub(r"\\s\+?", " ", pattern)
+    return re.sub(r"\s+", " ", pattern.replace("\\", ""))
+
+
+def test_route_specific_command_telemetry_is_weightless() -> None:
+    """A criterion that grades WHICH ROUTE ran must not move the score.
 
     THE GAP THIS EXISTS FOR. `pass_threshold: 0` was read as "this criterion is
     advisory", and the sibling test above enforces only that. It is half the
-    idiom. coder_eval's own field docs spell out the other half: "weight=0
-    excludes from the score but NOT from the pass/fail gate ... To make a
-    criterion truly non-gating, also set pass_threshold=0." A criterion with
-    `pass_threshold: 0` and `weight: 1.5` still lands in both halves of the
-    weighted mean, so failing it costs score without ever failing the task.
-
-    That is not neutral between arms, because a `command_executed` grades the
-    SHELL COMMAND an arm ran, and the two arms run different commands by
-    construction. In the 2026-09-23 same-ground run, 15 such criteria across
-    10 flow tasks scored 1.0 for v1 and 0.0 for v2 — `flow node add`,
-    `flow registry get`, `agent init --conversational`, `is connections list`
-    — dragging tasks that passed every graded check down to 0.55, 0.59, 0.71.
-    `datafabric_integration_create_get` returned SUCCESS at 0.55.
-
-    The outcome those criteria stand in for is graded by a `run_command`
-    against the artifact, which is route-blind. So the telemetry keeps
-    reporting and stops scoring.
-
-    Criteria carrying `stop_early` are exempt: there `weight` is load-bearing
-    for the pass-stop floor, not just for the score (see `ixp/routing.yaml`,
-    whose sentinel says so).
+    idiom; coder_eval's own field docs carry the other half: "weight=0 excludes
+    from the score but NOT from the pass/fail gate ... To make a criterion truly
+    non-gating, also set pass_threshold=0." So `pass_threshold: 0` with
+    `weight: 1.5` never fails a task and always moves its score.
+
+    That only matters where the command itself is route-specific. In the
+    2026-09-23 same-ground run those criteria scored 1.0 for v1 and 0.0 for v2,
+    dragging tasks that passed every graded check down with them —
+    `datafabric_integration_create_get` returned SUCCESS at 0.55 on three
+    `flow node add` advisories weighing 5.0 against two graded criteria at 3.0.
+
+    WHAT IS NOT IN `ROUTE_SPECIFIC_COMMANDS`, and why. An earlier revision of
+    this test keyed on criterion TYPE — every non-gating `command_executed` —
+    and that was wrong. It swept in `solution init`, `flow init`,
+    `flow validate`, `flow debug` and `flow eval ...`, which both routes run on
+    the same artifact and which the run shows both routes passing (31 of the 54
+    observed criteria were BOTH PASS). Zeroing those removes real, satisfiable
+    signal and buys no neutrality. `flow registry get|search|list` is out for
+    the same measured reason: the SDK arm passes those in the IxP tasks, so the
+    registry is not a v1-only surface — only `pull` is listed, as the refresh
+    step the SDK loop has no need of.
+
+    A criterion is therefore in scope only when its COMMAND has no counterpart
+    in the other route. Where an arm then fails one of the survivors, that is a
+    finding about the arm, which is the point.
+
+    `stop_early` criteria are exempt: there `weight` is load-bearing for the
+    pass-stop floor, not just for the score (see `ixp/routing.yaml`, whose
+    sentinel says so).
     """
     offenders = set()
     for relative, _, text in _tagged_tasks():
@@ -237,11 +261,14 @@ def test_non_gating_command_telemetry_is_weightless() -> None:
                 continue
             if re.search(r"(?m)^\s+stop_early:", criterion):
                 continue
+            pattern = _command_pattern(criterion)
+            if not any(re.search(family, pattern) for family in ROUTE_SPECIFIC_COMMANDS):
+                continue
             weight = re.search(r"(?m)^\s+weight:\s*([0-9.]+)", criterion)
             # An absent `weight` defaults to 1.0, so silence is not compliance.
             if weight is None or float(weight.group(1)) != 0:
                 offenders.add(relative)
-    assert offenders == NON_GATING_WEIGHT_ALLOWLIST
+    assert offenders == set()
 
 
 def test_gating_skill_telemetry_matches_the_temporary_allowlist() -> None:
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/generic_dynamic_node/generic_dynamic_node.yaml b/tests/tasks/uipath-maestro-flow/connector_features/generic_dynamic_node/generic_dynamic_node.yaml
index 7df3ce3aa4..4b2b88d77e 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/generic_dynamic_node/generic_dynamic_node.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/generic_dynamic_node/generic_dynamic_node.yaml
@@ -79,7 +79,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP|\$\{?UIP\}?)\s+(maestro\s+)?flow\s+debug'
     min_count: 1
-    weight: 0
+    weight: 1.5
     pass_threshold: 0.0
 
   # ── Execution: connector calls ServiceNow and surfaces an array output ──
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/jdbc_databricks_query/jdbc_databricks_query.yaml b/tests/tasks/uipath-maestro-flow/connector_features/jdbc_databricks_query/jdbc_databricks_query.yaml
index 2263dde29c..d53cf81b50 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/jdbc_databricks_query/jdbc_databricks_query.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/jdbc_databricks_query/jdbc_databricks_query.yaml
@@ -70,7 +70,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+(maestro\s+)?flow\s+debug'
     min_count: 1
-    weight: 0
+    weight: 1.5
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/connector_features/slack-http-fallback/slack_http_fallback.yaml b/tests/tasks/uipath-maestro-flow/connector_features/slack-http-fallback/slack_http_fallback.yaml
index 334b17e4ba..90562fe98c 100644
--- a/tests/tasks/uipath-maestro-flow/connector_features/slack-http-fallback/slack_http_fallback.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_features/slack-http-fallback/slack_http_fallback.yaml
@@ -56,7 +56,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: "(uip|\\$UIP)\\s+(maestro\\s+)?flow\\s+debug"
     min_count: 1
-    weight: 0
+    weight: 1.5
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/connector_trigger/webhook_waitfor_parallel.yaml b/tests/tasks/uipath-maestro-flow/connector_trigger/webhook_waitfor_parallel.yaml
index 0dc9a9d4fd..0821a6a271 100644
--- a/tests/tasks/uipath-maestro-flow/connector_trigger/webhook_waitfor_parallel.yaml
+++ b/tests/tasks/uipath-maestro-flow/connector_trigger/webhook_waitfor_parallel.yaml
@@ -69,7 +69,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+solution\s+(new|init)'
     min_count: 1
-    weight: 0
+    weight: 1.0
     pass_threshold: 0.0
 
   - type: run_command
@@ -85,7 +85,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+is\s+webhooks\s+config'
     min_count: 1
-    weight: 0
+    weight: 1.0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/conversational/conversational_chat_loop.yaml b/tests/tasks/uipath-maestro-flow/conversational/conversational_chat_loop.yaml
index 08dcbdf3f1..77e33ae70a 100644
--- a/tests/tasks/uipath-maestro-flow/conversational/conversational_chat_loop.yaml
+++ b/tests/tasks/uipath-maestro-flow/conversational/conversational_chat_loop.yaml
@@ -137,5 +137,5 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+.*--output\s+json'
     min_count: 1
-    weight: 0
+    weight: 1.0
     pass_threshold: 0
diff --git a/tests/tasks/uipath-maestro-flow/e2e/devcon_expense_approval.yaml b/tests/tasks/uipath-maestro-flow/e2e/devcon_expense_approval.yaml
index b12fd061fa..898f3639b6 100644
--- a/tests/tasks/uipath-maestro-flow/e2e/devcon_expense_approval.yaml
+++ b/tests/tasks/uipath-maestro-flow/e2e/devcon_expense_approval.yaml
@@ -56,7 +56,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP|\$\{?UIP\}?)\s+solution\s+(new|init)'
     min_count: 1
-    weight: 0
+    weight: 1.0
     pass_threshold: 0.0  # advisory: init auto-scaffolds the parent solution (CLI #2470)
 
   - type: command_executed
@@ -64,7 +64,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP|\$\{?UIP\}?)\s+(maestro\s+)?flow\s+init'
     min_count: 1
-    weight: 0
+    weight: 1.0
     pass_threshold: 0.0
 
   - type: run_command
@@ -88,7 +88,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP|\$\{?UIP\}?)\s+(maestro\s+)?flow\s+validate'
     min_count: 1
-    weight: 0
+    weight: 2.5
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/e2e/jira_lifecycle/jira_lifecycle.yaml b/tests/tasks/uipath-maestro-flow/e2e/jira_lifecycle/jira_lifecycle.yaml
index 7c4a858286..a58480e5c9 100644
--- a/tests/tasks/uipath-maestro-flow/e2e/jira_lifecycle/jira_lifecycle.yaml
+++ b/tests/tasks/uipath-maestro-flow/e2e/jira_lifecycle/jira_lifecycle.yaml
@@ -89,7 +89,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP|\$\{?UIP\}?)\s+(maestro\s+)?flow\s+debug'
     min_count: 1
-    weight: 0
+    weight: 1.5
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/evaluate/evaluator_type_choice.yaml b/tests/tasks/uipath-maestro-flow/evaluate/evaluator_type_choice.yaml
index 23ee88c538..f536ae4b9d 100644
--- a/tests/tasks/uipath-maestro-flow/evaluate/evaluator_type_choice.yaml
+++ b/tests/tasks/uipath-maestro-flow/evaluate/evaluator_type_choice.yaml
@@ -67,7 +67,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+solution\s+(new|init)'
     min_count: 1
-    weight: 0
+    weight: 1.0
     pass_threshold: 0.0  # advisory: init auto-scaffolds the parent solution (CLI #2470)
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/evaluate/local_crud.yaml b/tests/tasks/uipath-maestro-flow/evaluate/local_crud.yaml
index 680d107b75..ca3ea68415 100644
--- a/tests/tasks/uipath-maestro-flow/evaluate/local_crud.yaml
+++ b/tests/tasks/uipath-maestro-flow/evaluate/local_crud.yaml
@@ -58,7 +58,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+solution\s+(new|init)'
     min_count: 1
-    weight: 0
+    weight: 1.0
     pass_threshold: 0.0  # advisory: init auto-scaffolds the parent solution (CLI #2470)
 
   - type: command_executed
@@ -66,7 +66,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+(maestro\s+)?flow\s+init'
     min_count: 1
-    weight: 0
+    weight: 1.0
     pass_threshold: 0.0
 
   - type: run_command
@@ -106,7 +106,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+eval\s+(evaluator\s+list|set\s+list)'
     min_count: 1
-    weight: 0
+    weight: 1.0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/evaluate/no_auto_upload.yaml b/tests/tasks/uipath-maestro-flow/evaluate/no_auto_upload.yaml
index 77645c39f4..aadaa966eb 100644
--- a/tests/tasks/uipath-maestro-flow/evaluate/no_auto_upload.yaml
+++ b/tests/tasks/uipath-maestro-flow/evaluate/no_auto_upload.yaml
@@ -65,7 +65,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+solution\s+(new|init)'
     min_count: 1
-    weight: 0
+    weight: 1.0
     pass_threshold: 0.0  # advisory: init auto-scaffolds the parent solution (CLI #2470)
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/evaluate/simulation/simulation_crud.yaml b/tests/tasks/uipath-maestro-flow/evaluate/simulation/simulation_crud.yaml
index 4f23b449b5..1468ac2d3e 100644
--- a/tests/tasks/uipath-maestro-flow/evaluate/simulation/simulation_crud.yaml
+++ b/tests/tasks/uipath-maestro-flow/evaluate/simulation/simulation_crud.yaml
@@ -67,7 +67,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+solution\s+(new|init)'
     min_count: 1
-    weight: 0
+    weight: 1.0
     pass_threshold: 0.0
 
   - type: run_command
@@ -107,7 +107,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+eval\s+simulation\s+add\s+(?=.*--strategy\s+Static\b)(?=.*--mock-value\b)'
     min_count: 1
-    weight: 0
+    weight: 3.0
     pass_threshold: 0.0
 
   - type: run_command
@@ -123,7 +123,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+eval\s+simulation\s+list'
     min_count: 1
-    weight: 0
+    weight: 1.5
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/interactive/customer_escalation_triage/customer_escalation_triage.yaml b/tests/tasks/uipath-maestro-flow/interactive/customer_escalation_triage/customer_escalation_triage.yaml
index 551705ac81..db1ada8b25 100644
--- a/tests/tasks/uipath-maestro-flow/interactive/customer_escalation_triage/customer_escalation_triage.yaml
+++ b/tests/tasks/uipath-maestro-flow/interactive/customer_escalation_triage/customer_escalation_triage.yaml
@@ -92,7 +92,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP|\$\{?UIP\}?)\s+(maestro\s+)?flow\s+validate'
     min_count: 1
-    weight: 0
+    weight: 1.5
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/interactive/slack_channel_description_simulated/slack_channel_description_simulated.yaml b/tests/tasks/uipath-maestro-flow/interactive/slack_channel_description_simulated/slack_channel_description_simulated.yaml
index b30e7c331d..bfffd79a23 100644
--- a/tests/tasks/uipath-maestro-flow/interactive/slack_channel_description_simulated/slack_channel_description_simulated.yaml
+++ b/tests/tasks/uipath-maestro-flow/interactive/slack_channel_description_simulated/slack_channel_description_simulated.yaml
@@ -88,7 +88,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP|\$\{?UIP\}?)\s+(maestro\s+)?flow\s+debug'
     min_count: 1
-    weight: 0
+    weight: 1.5
     pass_threshold: 0.0
 
   # Runtime check (mirrors the retired non-simulated original, check_channel_description.py):
diff --git a/tests/tasks/uipath-maestro-flow/ixp/e2e_02_project_selection.yaml b/tests/tasks/uipath-maestro-flow/ixp/e2e_02_project_selection.yaml
index 5e1b045c25..6e86658646 100644
--- a/tests/tasks/uipath-maestro-flow/ixp/e2e_02_project_selection.yaml
+++ b/tests/tasks/uipath-maestro-flow/ixp/e2e_02_project_selection.yaml
@@ -77,7 +77,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: 'uip\s+maestro\s+flow\s+registry\s+(search|list|get)\b[^\n]*ixp'
     min_count: 1
-    weight: 0
+    weight: 2.0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/ixp/e2e_03_project_creation_handoff/e2e_03_project_creation_handoff.yaml b/tests/tasks/uipath-maestro-flow/ixp/e2e_03_project_creation_handoff/e2e_03_project_creation_handoff.yaml
index 4cf0a3edda..dd23318249 100644
--- a/tests/tasks/uipath-maestro-flow/ixp/e2e_03_project_creation_handoff/e2e_03_project_creation_handoff.yaml
+++ b/tests/tasks/uipath-maestro-flow/ixp/e2e_03_project_creation_handoff/e2e_03_project_creation_handoff.yaml
@@ -145,7 +145,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(?i)(uip|\$UIP)\s+maestro\s+flow\s+registry\s+search'
     min_count: 1
-    weight: 0
+    weight: 1.0
     pass_threshold: 0
 
   # Advisory until a real run proves it latches. impl.md requires the
@@ -157,5 +157,5 @@ success_criteria:
     description: "Advisory: handoff payload names the documents folder"
     command_pattern: 'uipath-ixp[\s\S]{0,600}documents'
     min_count: 1
-    weight: 0
+    weight: 1.0
     pass_threshold: 0
diff --git a/tests/tasks/uipath-maestro-flow/ixp/e2e_04_build_mechanics.yaml b/tests/tasks/uipath-maestro-flow/ixp/e2e_04_build_mechanics.yaml
index f5373060bb..1ad52ee475 100644
--- a/tests/tasks/uipath-maestro-flow/ixp/e2e_04_build_mechanics.yaml
+++ b/tests/tasks/uipath-maestro-flow/ixp/e2e_04_build_mechanics.yaml
@@ -123,7 +123,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(?i)(uip|\$UIP)\s+ixp\s+projects\s+create'
     min_count: 1
-    weight: 0
+    weight: 1.0
     pass_threshold: 0
 
   - type: command_executed
@@ -131,7 +131,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(?i)(uip|\$UIP)\s+ixp\s+deployments\s+create\b[^\n]*--folder-key'
     min_count: 1
-    weight: 0
+    weight: 1.0
     pass_threshold: 0
 
   # Primary gate: project created + folder-deployed this run, then the .flow
diff --git a/tests/tasks/uipath-maestro-flow/ixp/integration_handle_routing.yaml b/tests/tasks/uipath-maestro-flow/ixp/integration_handle_routing.yaml
index 807c7f4a28..6ff7381e77 100644
--- a/tests/tasks/uipath-maestro-flow/ixp/integration_handle_routing.yaml
+++ b/tests/tasks/uipath-maestro-flow/ixp/integration_handle_routing.yaml
@@ -69,7 +69,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: 'uip\s+maestro\s+flow\s+registry\s+(search|list|get)\b[^\n]*ixp'
     min_count: 1
-    weight: 0
+    weight: 2.0
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/ixp/routing_listing.yaml b/tests/tasks/uipath-maestro-flow/ixp/routing_listing.yaml
index 7ec546268d..4019ecf40e 100644
--- a/tests/tasks/uipath-maestro-flow/ixp/routing_listing.yaml
+++ b/tests/tasks/uipath-maestro-flow/ixp/routing_listing.yaml
@@ -65,7 +65,6 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(?i)uip\s+maestro\s+flow\s+registry\s+(list\b|(search|list)\b[^\n]*\b(ixp|document|extract))'
     min_count: 1
-    weight: 0
     pass_threshold: 0.0
 
   # A listing/Q&A task should not create a Flow artifact.
diff --git a/tests/tasks/uipath-maestro-flow/ixp/scaffold_minimal.yaml b/tests/tasks/uipath-maestro-flow/ixp/scaffold_minimal.yaml
index 0d30eff452..694d23148b 100644
--- a/tests/tasks/uipath-maestro-flow/ixp/scaffold_minimal.yaml
+++ b/tests/tasks/uipath-maestro-flow/ixp/scaffold_minimal.yaml
@@ -58,7 +58,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(?i)uip\s+maestro\s+flow\s+registry\s+(list\b|(search|list|get)\b[^\n]*\b(ixp|document|extract))'
     min_count: 1
-    weight: 0
+    weight: 2.0
     pass_threshold: 0.0
 
   - type: command_executed
@@ -66,7 +66,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: 'uip\s+(maestro\s+)?flow\s+validate'
     min_count: 1
-    weight: 0
+    weight: 1.5
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/ixp/scaffold_multinode.yaml b/tests/tasks/uipath-maestro-flow/ixp/scaffold_multinode.yaml
index 671b639c78..5213c520df 100644
--- a/tests/tasks/uipath-maestro-flow/ixp/scaffold_multinode.yaml
+++ b/tests/tasks/uipath-maestro-flow/ixp/scaffold_multinode.yaml
@@ -62,7 +62,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(?i)uip\s+maestro\s+flow\s+registry\s+(list\b|(search|list|get)\b[^\n]*\b(ixp|document|extract))'
     min_count: 1
-    weight: 0
+    weight: 2.0
     pass_threshold: 0.0
 
   - type: command_executed
@@ -70,7 +70,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: 'uip\s+(maestro\s+)?flow\s+validate'
     min_count: 1
-    weight: 0
+    weight: 1.5
     pass_threshold: 0.0
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/node_features/datafabric_native/integration_native_read_create.yaml b/tests/tasks/uipath-maestro-flow/node_features/datafabric_native/integration_native_read_create.yaml
index 66d5f1ebe3..ecda35c398 100644
--- a/tests/tasks/uipath-maestro-flow/node_features/datafabric_native/integration_native_read_create.yaml
+++ b/tests/tasks/uipath-maestro-flow/node_features/datafabric_native/integration_native_read_create.yaml
@@ -69,7 +69,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+maestro\s+flow\s+registry\s+get\s+"?core\.datafabric\.'
     min_count: 1
-    weight: 0
+    weight: 2.0
     pass_threshold: 0.0
 
   - type: command_executed
diff --git a/tests/tasks/uipath-maestro-flow/smoke/init_maestro_automate.yaml b/tests/tasks/uipath-maestro-flow/smoke/init_maestro_automate.yaml
index 8aa9a5073e..0660dec09d 100644
--- a/tests/tasks/uipath-maestro-flow/smoke/init_maestro_automate.yaml
+++ b/tests/tasks/uipath-maestro-flow/smoke/init_maestro_automate.yaml
@@ -35,7 +35,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+(maestro\s+)?flow\s+init\b(?=(?:(?!\n|&&|\|\||;|\||\s(?:uip|\$UIP)\s).)*--automate)'
     min_count: 1
-    weight: 0
+    weight: 1.5
     pass_threshold: 0.0
 
   # Scoped to the project the task asked for. A sandbox-wide search passes
diff --git a/tests/tasks/uipath-maestro-flow/smoke/init_validate.yaml b/tests/tasks/uipath-maestro-flow/smoke/init_validate.yaml
index 93efcc725e..fa1f91fa18 100644
--- a/tests/tasks/uipath-maestro-flow/smoke/init_validate.yaml
+++ b/tests/tasks/uipath-maestro-flow/smoke/init_validate.yaml
@@ -34,7 +34,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+solution\s+(init|new)'
     min_count: 1
-    weight: 0
+    weight: 1.5
     pass_threshold: 0.0  # advisory: init auto-scaffolds the parent solution (CLI #2470)
 
   - type: run_command
diff --git a/tests/tasks/uipath-maestro-flow/smoke/registry_discovery.yaml b/tests/tasks/uipath-maestro-flow/smoke/registry_discovery.yaml
index ef87d47738..40ec1b1b74 100644
--- a/tests/tasks/uipath-maestro-flow/smoke/registry_discovery.yaml
+++ b/tests/tasks/uipath-maestro-flow/smoke/registry_discovery.yaml
@@ -37,7 +37,7 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: 'uip\s+(maestro\s+)?flow\s+registry\s+pull'
     min_count: 1
-    weight: 1.0
+    weight: 0
     pass_threshold: 0.0
 
   - type: command_executed
diff --git a/tests/tasks/uipath-maestro-flow/voice/voice_outbound_call.yaml b/tests/tasks/uipath-maestro-flow/voice/voice_outbound_call.yaml
index dea9283907..25eca002c1 100644
--- a/tests/tasks/uipath-maestro-flow/voice/voice_outbound_call.yaml
+++ b/tests/tasks/uipath-maestro-flow/voice/voice_outbound_call.yaml
@@ -145,5 +145,5 @@ success_criteria:
     tool_name: "Bash"
     command_pattern: '(uip|\$UIP)\s+.*--output\s+json'
     min_count: 1
-    weight: 0
+    weight: 1.0
     pass_threshold: 0