#15 Codex GPT-5.6-sol-pro Ultra (Team Mode)

Quality

2026-07-12-codex-gpt-5-6-sol-ultra · Codex CLI v0.144.1

Orchestrator Tool Calls

817

8 unique tools

Sub-agent Tool Calls

1.4K

exec / apply_patch / read

Sub-agents

16

named specialist roles

Estimated Cost

$119.74

3.1M tokens

Tool Usage per Sub-agent

16 agents
Zeno /qa_engineer
358
Singer /frontend_architect
293
Tesla /frontend_architect/storefront_primitives
223
Sagan /backend_architect
82
Ptolemy /spec_auditor
74
Arendt /commerce_fixer
72
Herschel /security_reviewer
64
Anscombe /security_fixer
64
Confucius /qa_lead
38
Carson /qa_engineer/auth_export_qa
31
Ampere /livewire_runtime_fix
31
McClintock /tax_switch_fix
28
Beauvoir /qa_engineer/theme_search_qa
18
Dewey /customer_order_url_fix
15
Galileo /post_checkout_cart_fix
14
Harvey /backend_lead
6

Code Quality Assessment

PHPMetrics

Code Quality Assessment

PHPMetrics 2.9.1

Classes

286

Lines of Code

11.3K

Avg MI

70.5

Avg Complexity

6.8

Est. Bugs

51.2

max 3.0 in CheckRegistry

Violations

113

39 errors, 69 warnings

Complexity Distribution

Low (CCN <= 5)
202 (70.6%)
Medium (6-10)
29 (10.1%)
High (> 10)
55 (19.2%)

Maintainability Distribution

Excellent (>= 85)
45 (15.7%)
Good (70-84)
37 (12.9%)
Moderate (55-69)
119 (41.6%)
Low (< 55)
85 (29.7%)

Most Complex Classes

ProductService

App\Services

CCN 107

CheckRegistry

App\Services\CodeQuality

CCN 71

DiscountService

App\Services

CCN 58

Show

App\Livewire\Storefront\Checkout

CCN 50

OrderService

App\Services

CCN 48

Least Maintainable

ProductService

CCN 107 · 272 LOC

Show

CCN 50 · 297 LOC

CartService

CCN 36 · 183 LOC

CheckoutController

CCN 41 · 132 LOC

ProductController

CCN 30 · 116 LOC

Most Coupled Classes

AdminComponent

33

Order

31

Store

30

Product

27

User

27

Max Efferent

20

Max LCOM

7

Notes

  • •PHPMetrics flags 113 violations: 0 critical, 39 errors, 69 warnings, 5 info.
  • •SonarCloud reports 190 unresolved issues on PR #25: 34 critical, 147 major, 9 minor (50 bugs, 1 vulnerability, 139 code smells).
  • •Most complex class is App\Services\ProductService (CCN 107, MI 20.44) - by far the single worst-scoring class in the codebase.
  • •11.3K LOC across 286 classes - the leanest LOC of any build tested so far on this spec, despite the largest sub-agent team (16 named specialists).

SonarCloud Analysis

PR #25 - Quality Gate Failed
FAIL

Quality Gate: Failed

PR #25 · 2026-07-12-codex-gpt-5-6-sol-ultra → main

26.9h technical debt

190 issues total

View on SonarCloud ↗

Quality Gate Conditions

Reliability on New Code
C> 1
Security on New Code
C> 1
Maintainability on New Code
A> 1
Duplicated Lines on New Code
0.7%> 3.0%
Security Hotspots Reviewed
100.0%< 100%

Bugs

50

C

Reliability

Vulnerabilities

1

C

Security

Code Smells

139

A

Maintainability

Hotspots

0

100% reviewed

Lines of Code

18.9K

PHP

Duplication

0.8%

of all code

Coverage

0%

not configured

Issues by Severity

Critical
34
Major
147
Minor
9
Total190

Issues by Type

Code Smells
139 (73.2%)
Bugs
50 (26.3%)
Vulnerabilities
1 (0.5%)
Security Hotspots
0 (0.0%)
50 bugs (mostly minor logic-flow findings, not confirmed defects) and 1 vulnerability drag Reliability and Security down to C; Maintainability holds at A across 139 smells over 18.9K NCLOC.

Top Issue Rules

Table headers should have an id or scope attributeWeb:TableHeaderHasIdOrScopeCheck
39
MAJOR
Cognitive Complexity of methods should not be too highphp:S3776
23
CRITICAL
Input elements should have an autocomplete attributeWeb:S6840
19
MAJOR
Ternary operators should not be nestedphp:S3358
15
MAJOR
Unused function parameters should be removedphp:S1172
12
MAJOR
String literals should not be duplicatedphp:S1192
11
CRITICAL
Pages should have a titleWeb:PageWithoutTitleCheck
10
MAJOR
Landmarks should be used to improve accessibilityWeb:S6819
9
MAJOR
Methods should not contain too many return statementsphp:S1142
8
MAJOR
Generic exceptions should never be thrownphp:S112
8
MAJOR

Quality gate failed on two of five conditions: Reliability (C) and Security (C) on new code. Duplication (0.7%) and hotspot review (100%) both pass comfortably. The single biggest rule is missing table header id/scope attributes (39 hits) across admin data tables, followed by cognitive-complexity warnings concentrated almost entirely in one class, App\Services\ProductService (CCN 107) - the same class PHPMetrics independently flags as the codebase's worst offender.