diff --git a/.github/workflows/ecr-server.yml b/.github/workflows/ecr-server.yml
index ec504e75..085bd8ad 100644
--- a/.github/workflows/ecr-server.yml
+++ b/.github/workflows/ecr-server.yml
@@ -97,11 +97,12 @@ jobs:
           # headless, so host-skill discovery would put ~/.claude/skills in its
           # reach. Do not "fix" this back to "all" when the gate trips; read the
           # comment on the bundle before changing this value.
+          # The last column is the expected reasoning effort; None skips the check.
           EXPECT = {
-              "sei-spec":   ("none", 8, "speckit-"),
-              "root-cause": ("none", 1, None),
-              "xreview-scout-codex": ("none", 0, None),
-              "xreview-scout-cursor": ("none", 0, None),
+              "sei-spec":   ("none", 8, "speckit-", None),
+              "root-cause": ("none", 1, None, None),
+              "xreview-scout-codex": ("none", 0, None, "high"),
+              "xreview-scout-cursor": ("none", 0, None, None),
           }
           root = Path("/opt/sei-omnigent/agents")
           dirs = sorted(p for p in root.iterdir() if p.is_dir())
@@ -110,7 +111,7 @@ jobs:
           bad = False
           for d in dirs:
               spec = parse(d)
-              want_filter, want_skills, want_prefix = EXPECT[d.name]
+              want_filter, want_skills, want_prefix, want_effort = EXPECT[d.name]
               names = sorted(s.name for s in spec.skills)
               n = len(names)
               if spec.name != d.name:
@@ -123,6 +124,9 @@ jobs:
                   stray = [x for x in names if not x.startswith(want_prefix)]
                   if stray:
                       print(f"FAIL {d.name}: non-{want_prefix} skills vendored in: {stray}"); bad = True
+              effort = getattr(spec.executor, "reasoning_effort", None)
+              if want_effort is not None and effort != want_effort:
+                  print(f"FAIL {d.name}: reasoning effort {effort!r} != {want_effort!r}"); bad = True
               if spec.skills_filter == "none" and discover_host_skills(d, spec.skills_filter):
                   print(f"FAIL {d.name}: host scope leaked despite skills: none"); bad = True
               print(f"  ok {d.name} filter={spec.skills_filter!r} skills={n}")
diff --git a/.github/workflows/verify-agent-bundles.yml b/.github/workflows/verify-agent-bundles.yml
index 1419860b..0f217a6b 100644
--- a/.github/workflows/verify-agent-bundles.yml
+++ b/.github/workflows/verify-agent-bundles.yml
@@ -37,7 +37,7 @@ jobs:
       - name: Install the parser
         # No PyPI release matches the fork build the server runs, so this gate
         # approximates it. ecr-server.yml parses with what production has.
-        run: pip install --disable-pip-version-check 'omnigent==0.9.0'
+        run: pip install --disable-pip-version-check 'omnigent==0.16.0'
 
       - name: Parse every bundle and assert its contract
         run: |
@@ -55,7 +55,7 @@ jobs:
               # A scout carries no method and takes none from the host: the driver's
               # per-run prompt is its whole instruction, so a discovered skill would
               # be capability it was never asked to use.
-              'xreview-scout-codex': {'filter': 'none', 'skills': 0},
+              'xreview-scout-codex': {'filter': 'none', 'skills': 0, 'effort': 'high'},
               # Same contract on a different harness. Listed here because a bundle
               # with no entry fails below, and the deployment cannot reach its
               # harness yet: this gate is the only thing asserting it stays parsable
@@ -102,6 +102,12 @@ jobs:
                       print(f'FAIL {d.name}: non-{pref} skills vendored in: {stray}')
                       failed = True
 
+              # A parser too old to read the field reports None, which fails here too.
+              effort = getattr(spec.executor, 'reasoning_effort', None)
+              if 'effort' in exp and effort != exp['effort']:
+                  print(f'FAIL {d.name}: reasoning effort is {effort!r}, expected {exp["effort"]!r}')
+                  failed = True
+
               # With `none`, host discovery must return nothing even from a directory
               # that has host skills to offer.
               if spec.skills_filter == 'none':
diff --git a/agents/xreview-scout-codex/config.yaml b/agents/xreview-scout-codex/config.yaml
index 7151685f..dbdbef91 100644
--- a/agents/xreview-scout-codex/config.yaml
+++ b/agents/xreview-scout-codex/config.yaml
@@ -28,15 +28,12 @@ executor:
   # vendor-neutral gateway, read live off a process that was routed at
   # api.openai.com. Misreading it produced three wrong diagnoses.
   #
-  # gpt-5.6-sol is what ai-review's codex pass resolves to: that workflow pins no
-  # model and takes openai/codex-action's default, measured on a passing run.
-  # Matching it is the point — a scout on a different model than the reference
-  # reviewer is not the second opinion we mean to reproduce.
-  #
-  # Dotted, not dashed: codex spells it gpt-5.6-sol where the catalog writes
-  # gpt-5-6-sol. This value goes verbatim to `codex -c model=`, and the dotted
-  # form is the one the API accepts.
-  model: gpt-5.6-sol
+  # gpt-6.1-sol is OpenAI's balanced GPT-6 model: near-Astra on complex work at a
+  # lower cost, which suits one reading per pull request. Dotted, not dashed: this
+  # value goes verbatim to `codex -c model=`, and the dotted form is the one the
+  # API accepts.
+  model: gpt-6.1-sol
+  reasoning_effort: high
   config:
     # codex, and that is the point. The bundle is what fixes the harness,
     # so this is the one field that makes this a genuinely different reader