agentic-qe 3.10.5 → 3.10.7

package/assets/agents/v3/qe-risk-assessor.md

@@ -57,6 +57,14 @@ Proceed with analysis without confirmation when scope is clear.
 Apply impact analysis automatically for all risk assessments.
 Generate mitigation recommendations with cost-benefit analysis by default.
 </default_to_action>
+<evidence_discipline>
+ADR-105 evidence classes — label every finding you emit:
+- EXECUTED: you ran a real command; attach the command and its output as the artifact.
+- STATIC: derived from data (coverage file, AST, lockfile, schema); name the data source.
+- INFERRED: reasoning over code/content without execution. Never present it in the voice of verified fact.
+- CONJECTURE: pattern-matched heuristic or extrapolation; flag it as such.
+Quality gates block only on EXECUTED/STATIC; INFERRED routes to adversarial verification (ADR-102); CONJECTURE never gates. When a check can cheaply be executed instead of inferred, execute it and upgrade the label.
+</evidence_discipline>
 
 <parallel_execution>
 Assess risks across multiple components simultaneously.

package/assets/agents/v3/qe-root-cause-analyzer.md

@@ -68,6 +68,14 @@ Proceed with investigation without confirmation when artifacts are available.
 Apply pattern correlation automatically across related failures.
 Generate prevention recommendations by default for all root causes.
 </default_to_action>
+<evidence_discipline>
+ADR-105 evidence classes — label every finding you emit:
+- EXECUTED: you ran a real command; attach the command and its output as the artifact.
+- STATIC: derived from data (coverage file, AST, lockfile, schema); name the data source.
+- INFERRED: reasoning over code/content without execution. Never present it in the voice of verified fact.
+- CONJECTURE: pattern-matched heuristic or extrapolation; flag it as such.
+Quality gates block only on EXECUTED/STATIC; INFERRED routes to adversarial verification (ADR-102); CONJECTURE never gates. When a check can cheaply be executed instead of inferred, execute it and upgrade the label.
+</evidence_discipline>
 
 <parallel_execution>
 Analyze multiple failures simultaneously.

package/assets/agents/v3/qe-sap-idoc-tester.md

@@ -46,6 +46,14 @@ Apply strict field validation for production IDoc types, relaxed for development
 Use assertEventually with 30-second default timeout and 2-second polling interval for async processing.
 Automatically detect IDoc basic type from message type when not explicitly specified.
 </default_to_action>
+<evidence_discipline>
+ADR-105 evidence classes — label every finding you emit:
+- EXECUTED: you ran a real command; attach the command and its output as the artifact.
+- STATIC: derived from data (coverage file, AST, lockfile, schema); name the data source.
+- INFERRED: reasoning over code/content without execution. Never present it in the voice of verified fact.
+- CONJECTURE: pattern-matched heuristic or extrapolation; flag it as such.
+Quality gates block only on EXECUTED/STATIC; INFERRED routes to adversarial verification (ADR-102); CONJECTURE never gates. When a check can cheaply be executed instead of inferred, execute it and upgrade the label.
+</evidence_discipline>
 
 <parallel_execution>
 Validate multiple IDoc types simultaneously across different message types.

package/assets/agents/v3/qe-sap-rfc-tester.md

@@ -45,6 +45,14 @@ Apply strict type checking against ABAP data dictionary types (CHAR, NUMC, DATS,
 Test BAPI return structures (TYPE, ID, NUMBER, MESSAGE) for all error paths.
 Use BAPI_TRANSACTION_ROLLBACK after every test to avoid persistent data changes.
 </default_to_action>
+<evidence_discipline>
+ADR-105 evidence classes — label every finding you emit:
+- EXECUTED: you ran a real command; attach the command and its output as the artifact.
+- STATIC: derived from data (coverage file, AST, lockfile, schema); name the data source.
+- INFERRED: reasoning over code/content without execution. Never present it in the voice of verified fact.
+- CONJECTURE: pattern-matched heuristic or extrapolation; flag it as such.
+Quality gates block only on EXECUTED/STATIC; INFERRED routes to adversarial verification (ADR-102); CONJECTURE never gates. When a check can cheaply be executed instead of inferred, execute it and upgrade the label.
+</evidence_discipline>
 
 <parallel_execution>
 Test multiple BAPIs/RFCs simultaneously across connection pool.

package/assets/agents/v3/qe-security-auditor.md

@@ -63,6 +63,14 @@ When auditing credential files (.env, .env.*, secrets), ALWAYS check .gitignore
 - Files NOT in .gitignore: report as CRITICAL (secrets committed to version control).
 - Hardcoded secrets in source code (.ts, .js, etc.) are always CRITICAL regardless of .gitignore.
 </default_to_action>
+<evidence_discipline>
+ADR-105 evidence classes — label every finding you emit:
+- EXECUTED: you ran a real command; attach the command and its output as the artifact.
+- STATIC: derived from data (coverage file, AST, lockfile, schema); name the data source.
+- INFERRED: reasoning over code/content without execution. Never present it in the voice of verified fact.
+- CONJECTURE: pattern-matched heuristic or extrapolation; flag it as such.
+Quality gates block only on EXECUTED/STATIC; INFERRED routes to adversarial verification (ADR-102); CONJECTURE never gates. When a check can cheaply be executed instead of inferred, execute it and upgrade the label.
+</evidence_discipline>
 
 <parallel_execution>
 Audit multiple security categories simultaneously.

package/assets/agents/v3/qe-security-scanner.md

@@ -50,6 +50,14 @@ Proceed with scanning without confirmation when scope is clear.
 Apply all relevant rule sets automatically based on detected language/framework.
 Use incremental scanning for known codebases to reduce scan time.
 </default_to_action>
+<evidence_discipline>
+ADR-105 evidence classes — label every finding you emit:
+- EXECUTED: you ran a real command; attach the command and its output as the artifact.
+- STATIC: derived from data (coverage file, AST, lockfile, schema); name the data source.
+- INFERRED: reasoning over code/content without execution. Never present it in the voice of verified fact.
+- CONJECTURE: pattern-matched heuristic or extrapolation; flag it as such.
+Quality gates block only on EXECUTED/STATIC; INFERRED routes to adversarial verification (ADR-102); CONJECTURE never gates. When a check can cheaply be executed instead of inferred, execute it and upgrade the label.
+</evidence_discipline>
 
 <parallel_execution>
 Run SAST, dependency, and secrets scans in parallel.

package/assets/agents/v3/qe-soap-tester.md

@@ -44,6 +44,14 @@ Test WS-Security configurations automatically when security headers are detected
 Validate SOAP fault codes and fault string content for all error paths.
 Use SOAP 1.2 by default unless SOAP 1.1 binding is explicitly declared.
 </default_to_action>
+<evidence_discipline>
+ADR-105 evidence classes — label every finding you emit:
+- EXECUTED: you ran a real command; attach the command and its output as the artifact.
+- STATIC: derived from data (coverage file, AST, lockfile, schema); name the data source.
+- INFERRED: reasoning over code/content without execution. Never present it in the voice of verified fact.
+- CONJECTURE: pattern-matched heuristic or extrapolation; flag it as such.
+Quality gates block only on EXECUTED/STATIC; INFERRED routes to adversarial verification (ADR-102); CONJECTURE never gates. When a check can cheaply be executed instead of inferred, execute it and upgrade the label.
+</evidence_discipline>
 
 <parallel_execution>
 Parse multiple WSDL documents simultaneously for composite service testing.

package/assets/agents/v3/qe-sod-analyzer.md

@@ -47,6 +47,14 @@ Automatically detect authorization object types and applicable conflict rules.
 Flag any role with both "create" and "approve" activities on the same business object as HIGH risk by default.
 Generate audit documentation in parallel with conflict analysis.
 </default_to_action>
+<evidence_discipline>
+ADR-105 evidence classes — label every finding you emit:
+- EXECUTED: you ran a real command; attach the command and its output as the artifact.
+- STATIC: derived from data (coverage file, AST, lockfile, schema); name the data source.
+- INFERRED: reasoning over code/content without execution. Never present it in the voice of verified fact.
+- CONJECTURE: pattern-matched heuristic or extrapolation; flag it as such.
+Quality gates block only on EXECUTED/STATIC; INFERRED routes to adversarial verification (ADR-102); CONJECTURE never gates. When a check can cheaply be executed instead of inferred, execute it and upgrade the label.
+</evidence_discipline>
 
 <parallel_execution>
 Analyze multiple roles for SoD conflicts simultaneously.

package/assets/agents/v3/qe-tdd-specialist.md

@@ -40,6 +40,14 @@ Proceed through RED-GREEN-REFACTOR without confirmation for clear requirements.
 Apply London or Chicago school based on code context automatically.
 Generate minimal implementation guidance during GREEN phase.
 </default_to_action>
+<evidence_discipline>
+ADR-105 evidence classes — label every finding you emit:
+- EXECUTED: you ran a real command; attach the command and its output as the artifact.
+- STATIC: derived from data (coverage file, AST, lockfile, schema); name the data source.
+- INFERRED: reasoning over code/content without execution. Never present it in the voice of verified fact.
+- CONJECTURE: pattern-matched heuristic or extrapolation; flag it as such.
+Quality gates block only on EXECUTED/STATIC; INFERRED routes to adversarial verification (ADR-102); CONJECTURE never gates. When a check can cheaply be executed instead of inferred, execute it and upgrade the label.
+</evidence_discipline>
 
 <parallel_execution>
 Execute multiple TDD cycles for independent features simultaneously.

package/assets/agents/v3/qe-test-architect.md

@@ -76,6 +76,14 @@ Proceed with test creation without asking for confirmation when framework and ta
 Apply learned patterns automatically based on code analysis and past experience.
 Use the test pyramid principle: 70% unit, 20% integration, 10% e2e.
 </default_to_action>
+<evidence_discipline>
+ADR-105 evidence classes — label every finding you emit:
+- EXECUTED: you ran a real command; attach the command and its output as the artifact.
+- STATIC: derived from data (coverage file, AST, lockfile, schema); name the data source.
+- INFERRED: reasoning over code/content without execution. Never present it in the voice of verified fact.
+- CONJECTURE: pattern-matched heuristic or extrapolation; flag it as such.
+Quality gates block only on EXECUTED/STATIC; INFERRED routes to adversarial verification (ADR-102); CONJECTURE never gates. When a check can cheaply be executed instead of inferred, execute it and upgrade the label.
+</evidence_discipline>
 
 <parallel_execution>
 Analyze multiple source files simultaneously for faster test planning.

package/assets/agents/v3/qe-test-idea-rewriter.md

@@ -38,6 +38,14 @@ Preserve all metadata (IDs, priorities, automation types).
 Process entire assessment files in single pass.
 Validate zero "Verify" patterns remain post-transformation.
 </default_to_action>
+<evidence_discipline>
+ADR-105 evidence classes — label every finding you emit:
+- EXECUTED: you ran a real command; attach the command and its output as the artifact.
+- STATIC: derived from data (coverage file, AST, lockfile, schema); name the data source.
+- INFERRED: reasoning over code/content without execution. Never present it in the voice of verified fact.
+- CONJECTURE: pattern-matched heuristic or extrapolation; flag it as such.
+Quality gates block only on EXECUTED/STATIC; INFERRED routes to adversarial verification (ADR-102); CONJECTURE never gates. When a check can cheaply be executed instead of inferred, execute it and upgrade the label.
+</evidence_discipline>
 
 <parallel_execution>
 Process multiple test ideas simultaneously.

package/assets/agents/v3/qe-transfer-specialist.md

@@ -38,6 +38,14 @@ Proceed with cross-framework mapping without confirmation when mappings are avai
 Apply negative transfer prevention automatically during all transfers.
 Generate transfer compatibility reports by default for new agent pairs.
 </default_to_action>
+<evidence_discipline>
+ADR-105 evidence classes — label every finding you emit:
+- EXECUTED: you ran a real command; attach the command and its output as the artifact.
+- STATIC: derived from data (coverage file, AST, lockfile, schema); name the data source.
+- INFERRED: reasoning over code/content without execution. Never present it in the voice of verified fact.
+- CONJECTURE: pattern-matched heuristic or extrapolation; flag it as such.
+Quality gates block only on EXECUTED/STATIC; INFERRED routes to adversarial verification (ADR-102); CONJECTURE never gates. When a check can cheaply be executed instead of inferred, execute it and upgrade the label.
+</evidence_discipline>
 
 <parallel_execution>
 Transfer knowledge across multiple agent pairs simultaneously.

package/assets/agents/v3/qe-visual-tester.md

@@ -39,6 +39,14 @@ Proceed with testing without confirmation when baselines exist.
 Apply AI comparison for semantic changes automatically.
 Use multi-viewport testing by default for responsive components.
 </default_to_action>
+<evidence_discipline>
+ADR-105 evidence classes — label every finding you emit:
+- EXECUTED: you ran a real command; attach the command and its output as the artifact.
+- STATIC: derived from data (coverage file, AST, lockfile, schema); name the data source.
+- INFERRED: reasoning over code/content without execution. Never present it in the voice of verified fact.
+- CONJECTURE: pattern-matched heuristic or extrapolation; flag it as such.
+Quality gates block only on EXECUTED/STATIC; INFERRED routes to adversarial verification (ADR-102); CONJECTURE never gates. When a check can cheaply be executed instead of inferred, execute it and upgrade the label.
+</evidence_discipline>
 
 <parallel_execution>
 Capture screenshots across multiple viewports simultaneously.

package/assets/skills/.validation/examples/chaos-engineering-output.example.json

@@ -500,7 +500,7 @@
     "executionTimeMs": 2160000,
     "toolsUsed": ["toxiproxy", "kubectl", "stress-ng", "custom"],
     "agentId": "qe-chaos-engineer",
-    "modelUsed": "claude-3.5-sonnet",
+    "modelUsed": "claude-sonnet-4-6",
     "environment": "staging"
   },
   "validation": {

package/assets/skills/.validation/examples/performance-testing-output.example.json

@@ -230,7 +230,7 @@
     "executionTimeMs": 305000,
     "toolsUsed": ["k6", "prometheus", "grafana"],
     "agentId": "qe-performance-tester",
-    "modelUsed": "claude-3.5-sonnet",
+    "modelUsed": "claude-sonnet-4-6",
     "targetUrl": "https://api.example.com",
     "environment": "staging"
   },

package/assets/skills/.validation/examples/security-testing-output.example.json

@@ -380,7 +380,7 @@
     "executionTimeMs": 45230,
     "toolsUsed": ["semgrep", "npm-audit", "owasp-zap", "trivy"],
     "agentId": "qe-security-scanner",
-    "modelUsed": "claude-3.5-sonnet",
+    "modelUsed": "claude-sonnet-4-6",
     "inputHash": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855",
     "targetUrl": "https://example.com",
     "targetPath": "src/",

package/assets/skills/.validation/examples/testability-scoring-output.example.json

@@ -318,7 +318,7 @@
     "executionTimeMs": 12450,
     "toolsUsed": ["playwright", "chrome-devtools", "axe-core"],
     "agentId": "qe-quality-analyzer",
-    "modelUsed": "claude-3.5-sonnet",
+    "modelUsed": "claude-sonnet-4-6",
     "inputHash": "7d865e959b2466918c9863afca942d0fb89d7c9ac0c99bafc3749504ded97730",
     "targetUrl": "https://example.com",
     "environment": "ci",

package/assets/skills/.validation/schemas/skill-eval.schema.json

@@ -26,19 +26,17 @@
       "items": {
         "type": "string",
         "enum": [
+          "claude-opus-4-8",
           "claude-opus-4-7",
           "claude-sonnet-4-6",
           "claude-haiku-4-5",
           "claude-opus-4-5",
-          "claude-sonnet-4",
-          "claude-3.5-sonnet",
-          "claude-3-haiku",
           "gpt-4o",
           "gpt-4o-mini",
           "gpt-4-turbo"
         ]
       },
-      "default": ["claude-3.5-sonnet"],
+      "default": ["claude-sonnet-4-6"],
       "minItems": 1,
       "uniqueItems": true,
       "description": "Models to run evaluation against for cross-model validation"

package/assets/skills/.validation/schemas/skill-frontmatter.schema.json

@@ -264,7 +264,7 @@
         "validation_models": {
           "type": "array",
           "items": { "type": "string" },
-          "description": "Models used for evaluation testing (e.g., ['claude-3.5-sonnet', 'claude-3-haiku'])"
+          "description": "Models used for evaluation testing (e.g., ['claude-sonnet-4-6', 'claude-haiku-4-5'])"
         },
         "pass_rate": {
           "type": "number",

package/assets/skills/.validation/schemas/skill-output.template.json

@@ -120,7 +120,7 @@
         },
         "modelUsed": {
           "type": "string",
-          "description": "LLM model used for execution (e.g., claude-3.5-sonnet)"
+          "description": "LLM model used for execution (e.g., claude-sonnet-4-6)"
         },
         "inputHash": {
           "type": "string",

package/assets/skills/.validation/templates/eval.template.yaml

@@ -31,9 +31,9 @@ description: >
 # model-specific quirks. Results are compared to detect variance.
 
 models_to_test:
-  - claude-3.5-sonnet    # Primary model (high accuracy expected)
-  - claude-3-haiku       # Fast model (ensure it meets minimum quality)
-  # - gpt-4o             # Optional: Cross-vendor validation
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
+  # - gpt-4o             # Optional: cross-vendor validation
 
 # =============================================================================
 # MCP Integration Configuration (NEW in v1.4)

package/assets/skills/.validation/templates/security-testing-eval.template.yaml

@@ -32,8 +32,9 @@ description: >
 # model-specific quirks. Results are compared to detect variance.
 
 models_to_test:
-  - claude-3.5-sonnet    # Primary model (high accuracy expected)
-  - claude-3-haiku       # Fast model (ensure it meets minimum quality)
+  - claude-opus-4-8      # Capability ceiling (high-stakes skill)
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
   - gpt-4o               # Cross-vendor validation (optional)
 
 # =============================================================================

package/assets/skills/.validation/templates/skill-frontmatter.example.yaml

@@ -202,8 +202,8 @@ validation:
 
   # Multi-model testing configuration
   validation_models:
-    - claude-3.5-sonnet
-    - claude-3-haiku
+    - claude-sonnet-4-6
+    - claude-haiku-4-5
     - gpt-4o
 
   # Evaluation metrics

package/assets/skills/a11y-ally/evals/a11y-ally.yaml

@@ -23,8 +23,8 @@ description: >
 # =============================================================================
 
 models_to_test:
-  - claude-3.5-sonnet
-  - claude-3-haiku
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 # =============================================================================
 # MCP Integration Configuration

package/assets/skills/accessibility-testing/evals/accessibility-testing.yaml

@@ -27,9 +27,8 @@ description: >
 # =============================================================================
 
 models_to_test:
-  - claude-3.5-sonnet    # Primary model (high accuracy expected)
-  - claude-haiku-4-5          # Fast model (ensure minimum quality)
-  - gpt-4o               # Cross-vendor validation
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 # =============================================================================
 # MCP Integration Configuration

package/assets/skills/api-testing-patterns/evals/api-testing-patterns.yaml

@@ -30,9 +30,8 @@ description: >
 # =============================================================================
 
 models_to_test:
-  - claude-sonnet-4-6        # Primary model (high accuracy expected)
-  - claude-haiku-4-5          # Fast model (minimum quality bar)
-  - gpt-4o                # Cross-vendor validation
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 # =============================================================================
 # MCP Integration Configuration

package/assets/skills/chaos-engineering-resilience/evals/chaos-engineering-resilience.yaml

@@ -34,9 +34,9 @@ description: >
 # =============================================================================
 
 models_to_test:
-  - claude-3.5-sonnet    # Primary model (high accuracy expected)
-  - claude-haiku-4-5          # Fast model (minimum quality threshold)
-  - gpt-4o               # Cross-vendor validation
+  - claude-opus-4-8      # Capability ceiling (high-stakes skill)
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 # =============================================================================
 # MCP Integration Configuration

package/assets/skills/cicd-pipeline-qe-orchestrator/evals/cicd-pipeline-qe-orchestrator.yaml

@@ -6,8 +6,8 @@ description: >
   and quality gate configuration across multiple models.
 
 models_to_test:
-  - claude-3.5-sonnet
-  - claude-3-haiku
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 mcp_integration:
   enabled: true

package/assets/skills/compatibility-testing/evals/compatibility-testing.yaml

@@ -15,8 +15,8 @@ description: >
 # =============================================================================
 
 models_to_test:
-  - claude-3.5-sonnet
-  - claude-3-haiku
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 # =============================================================================
 # MCP Integration Configuration

package/assets/skills/compliance-testing/evals/compliance-testing.yaml

@@ -31,9 +31,8 @@ description: >
 # =============================================================================
 
 models_to_test:
-  - claude-3.5-sonnet    # Primary model (high accuracy expected)
-  - claude-3-haiku       # Fast model (minimum quality threshold)
-  - gpt-4o               # Cross-vendor validation
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 # =============================================================================
 # MCP Integration Configuration

package/assets/skills/contract-testing/evals/contract-testing.yaml

@@ -30,9 +30,8 @@ description: >
 # =============================================================================
 
 models_to_test:
-  - claude-sonnet-4-6        # Primary model (high accuracy expected)
-  - claude-haiku-4-5          # Fast model (minimum quality bar)
-  - gpt-4o                # Cross-vendor validation
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 # =============================================================================
 # MCP Integration Configuration

package/assets/skills/database-testing/evals/database-testing.yaml

@@ -33,9 +33,8 @@ description: >
 # =============================================================================
 
 models_to_test:
-  - claude-sonnet-4-6      # Primary model (high accuracy expected)
-  - claude-haiku-4-5         # Fast model (minimum quality threshold)
-  - gpt-4o               # Cross-vendor validation
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 # =============================================================================
 # MCP Integration Configuration

package/assets/skills/enterprise-integration-testing/evals/enterprise-integration-testing.yaml

@@ -6,8 +6,8 @@ description: >
   data validation, middleware testing, and enterprise quality gates.
 
 models_to_test:
-  - claude-3.5-sonnet
-  - claude-3-haiku
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 mcp_integration:
   enabled: true

package/assets/skills/localization-testing/evals/localization-testing.yaml

@@ -31,8 +31,8 @@ description: >
 # =============================================================================
 
 models_to_test:
-  - claude-3.5-sonnet    # Primary model (high accuracy expected)
-  - claude-haiku-4-5          # Fast model (minimum quality threshold)
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 # =============================================================================
 # MCP Integration Configuration

package/assets/skills/middleware-testing-patterns/evals/middleware-testing-patterns.yaml

@@ -6,8 +6,8 @@ description: >
   ESB error handling, and Enterprise Integration Patterns.
 
 models_to_test:
-  - claude-3.5-sonnet
-  - claude-3-haiku
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 mcp_integration:
   enabled: true

package/assets/skills/mobile-testing/evals/mobile-testing.yaml

@@ -31,8 +31,8 @@ description: >
 # =============================================================================
 
 models_to_test:
-  - claude-3.5-sonnet    # Primary model
-  - claude-3-haiku       # Fast model (minimum quality)
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 # =============================================================================
 # MCP Integration Configuration

package/assets/skills/mutation-testing/evals/mutation-testing.yaml

@@ -25,9 +25,8 @@ description: >
 # =============================================================================
 
 models_to_test:
-  - claude-3.5-sonnet    # Primary model (high accuracy expected)
-  - claude-3-haiku       # Fast model (ensure minimum quality)
-  - gpt-4o               # Cross-vendor validation
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 # =============================================================================
 # MCP Integration Configuration

package/assets/skills/mutation-testing/test-data/sample-output.json

@@ -272,7 +272,7 @@
     "executionTimeMs": 125430,
     "toolsUsed": ["stryker"],
     "agentId": "qe-mutation-tester",
-    "modelUsed": "claude-3.5-sonnet",
+    "modelUsed": "claude-sonnet-4-6",
     "targetPath": "src/auth",
     "testPath": "tests/auth",
     "incremental": false,

package/assets/skills/n8n-expression-testing/evals/n8n-expression-testing.yaml

@@ -32,8 +32,8 @@ description: >
 # =============================================================================
 
 models_to_test:
-  - claude-3.5-sonnet
-  - claude-3-haiku
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 # =============================================================================
 # MCP Integration Configuration

package/assets/skills/n8n-integration-testing-patterns/evals/n8n-integration-testing-patterns.yaml

@@ -32,8 +32,8 @@ description: >
 # =============================================================================
 
 models_to_test:
-  - claude-3.5-sonnet
-  - claude-3-haiku
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 # =============================================================================
 # MCP Integration Configuration

package/assets/skills/n8n-security-testing/evals/n8n-security-testing.yaml

@@ -32,8 +32,9 @@ description: >
 # =============================================================================
 
 models_to_test:
-  - claude-3.5-sonnet
-  - claude-3-haiku
+  - claude-opus-4-8      # Capability ceiling (high-stakes skill)
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 # =============================================================================
 # MCP Integration Configuration

package/assets/skills/n8n-trigger-testing-strategies/evals/n8n-trigger-testing-strategies.yaml

@@ -31,8 +31,8 @@ description: >
 # =============================================================================
 
 models_to_test:
-  - claude-3.5-sonnet
-  - claude-haiku-4-5
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 # =============================================================================
 # MCP Integration Configuration

package/assets/skills/n8n-workflow-testing-fundamentals/evals/n8n-workflow-testing-fundamentals.yaml

@@ -32,8 +32,8 @@ description: >
 # =============================================================================
 
 models_to_test:
-  - claude-3.5-sonnet
-  - claude-3-haiku
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 # =============================================================================
 # MCP Integration Configuration

package/assets/skills/observability-testing-patterns/evals/observability-testing-patterns.yaml

@@ -6,8 +6,8 @@ description: >
   and SLA/SLO verification.
 
 models_to_test:
-  - claude-3.5-sonnet
-  - claude-3-haiku
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 mcp_integration:
   enabled: true

package/assets/skills/pentest-validation/evals/pentest-validation.yaml

@@ -32,9 +32,9 @@ description: >
 # =============================================================================
 
 models_to_test:
-  - claude-3.5-sonnet    # Primary model (high accuracy expected)
-  - claude-3-haiku       # Fast model (minimum quality threshold)
-  - gpt-4o               # Cross-vendor validation
+  - claude-opus-4-8      # Capability ceiling (high-stakes skill)
+  - claude-sonnet-4-6    # Primary (high accuracy expected)
+  - claude-haiku-4-5     # Fast model (minimum quality floor)
 
 # =============================================================================
 # MCP Integration Configuration