diff --git a/reports/architecture_maintainability.json b/reports/architecture_maintainability.json
index c9c3dd0..33012b9 100644
--- a/reports/architecture_maintainability.json
+++ b/reports/architecture_maintainability.json
@@ -4,10 +4,10 @@
   "generated_at": "2026-06-16",
   "skill_dir": ".",
   "summary": {
-    "python_file_count": 211,
-    "script_file_count": 137,
+    "python_file_count": 212,
+    "script_file_count": 138,
     "test_file_count": 74,
-    "internal_module_count": 53,
+    "internal_module_count": 54,
     "cli_script_count": 86,
     "command_handler_count": 68,
     "entrypoint_command_handler_count": 18,
@@ -37,15 +37,15 @@
       "recommendation": "Split viewer data assembly from HTML section rendering."
     },
     {
-      "path": "scripts/skill_report_model.py",
-      "lines": 665,
-      "kind": "internal-module",
+      "path": "tests/verify_world_class_evidence_intake.py",
+      "lines": 677,
+      "kind": "test",
       "severity": "pass",
-      "recommendation": "Watch this file before adding new responsibilities; extract a helper module when one concern dominates."
+      "recommendation": "Break broad integration assertions into focused verifier helpers when the next behavior change lands."
     },
     {
-      "path": "scripts/world_class_evidence_contract.py",
-      "lines": 654,
+      "path": "scripts/skill_report_model.py",
+      "lines": 665,
       "kind": "internal-module",
       "severity": "pass",
       "recommendation": "Watch this file before adding new responsibilities; extract a helper module when one concern dominates."
@@ -93,11 +93,11 @@
       "recommendation": "Watch this file before adding new responsibilities; extract a helper module when one concern dominates."
     },
     {
-      "path": "tests/verify_world_class_evidence_intake.py",
-      "lines": 611,
-      "kind": "test",
+      "path": "scripts/world_class_evidence_contract.py",
+      "lines": 618,
+      "kind": "internal-module",
       "severity": "pass",
-      "recommendation": "Break broad integration assertions into focused verifier helpers when the next behavior change lands."
+      "recommendation": "Watch this file before adding new responsibilities; extract a helper module when one concern dominates."
     },
     {
       "path": "scripts/render_benchmark_reproducibility.py",
diff --git a/reports/architecture_maintainability.md b/reports/architecture_maintainability.md
index 7b75ccc..4e0fc49 100644
--- a/reports/architecture_maintainability.md
+++ b/reports/architecture_maintainability.md
@@ -5,10 +5,10 @@ Generated at: `2026-06-16`
 ## Summary
 
 - decision: `pass`
-- python files: `211`
-- scripts: `137`
+- python files: `212`
+- scripts: `138`
 - tests: `74`
-- internal modules: `53`
+- internal modules: `54`
 - CLI scripts: `86`
 - Yao CLI command handlers: `68`
 - entrypoint command handlers: `18`
@@ -35,15 +35,15 @@ No near-threshold files found.
 | --- | ---: | --- | --- |
 | `tests/verify_yao_cli.py` | `696` | `test` | `pass` |
 | `scripts/render_review_viewer.py` | `685` | `cli-script` | `pass` |
+| `tests/verify_world_class_evidence_intake.py` | `677` | `test` | `pass` |
 | `scripts/skill_report_model.py` | `665` | `internal-module` | `pass` |
-| `scripts/world_class_evidence_contract.py` | `654` | `internal-module` | `pass` |
 | `scripts/render_skill_os2_coverage.py` | `649` | `cli-script` | `pass` |
 | `scripts/render_review_studio.py` | `647` | `cli-script` | `pass` |
 | `scripts/render_reference_synthesis.py` | `644` | `cli-script` | `pass` |
 | `scripts/review_studio_gates.py` | `643` | `internal-module` | `pass` |
 | `scripts/cross_packager.py` | `638` | `cli-script` | `pass` |
 | `scripts/build_skill_atlas.py` | `637` | `cli-script` | `pass` |
-| `tests/verify_world_class_evidence_intake.py` | `611` | `test` | `pass` |
+| `scripts/world_class_evidence_contract.py` | `618` | `internal-module` | `pass` |
 | `scripts/render_benchmark_reproducibility.py` | `595` | `cli-script` | `pass` |
 
 ## Release Rule
diff --git a/reports/context_budget.json b/reports/context_budget.json
index bf3c866..1ac6299 100644
--- a/reports/context_budget.json
+++ b/reports/context_budget.json
@@ -6,16 +6,16 @@
     "context_budget_tier": "production",
     "context_budget_limit": 1000,
     "skill_body_tokens": 797,
-    "other_text_tokens": 1064742,
+    "other_text_tokens": 1067291,
     "estimated_initial_load_tokens": 990,
-    "estimated_total_text_tokens": 1065539,
-    "deferred_resource_tokens": 493019,
+    "estimated_total_text_tokens": 1068088,
+    "deferred_resource_tokens": 494370,
     "deferred_resource_warn_threshold": 120000,
     "deferred_resource_dirs": [
       {
         "path": "scripts",
-        "estimated_tokens": 432921,
-        "file_count": 137
+        "estimated_tokens": 434272,
+        "file_count": 138
       },
       {
         "path": "references",
@@ -36,8 +36,8 @@
     "large_deferred_resource_dirs": [
       {
         "path": "scripts",
-        "estimated_tokens": 432921,
-        "file_count": 137
+        "estimated_tokens": 434272,
+        "file_count": 138
       }
     ],
     "deferred_resource_governance": {
@@ -59,14 +59,14 @@
           ],
           "missing": [],
           "path": "scripts",
-          "estimated_tokens": 432921,
-          "file_count": 137,
+          "estimated_tokens": 434272,
+          "file_count": 138,
           "rationale": "Script resources are deterministic deferred tools, not initial-load prompt context."
         }
       ],
       "summary": "Large deferred resources are indexed and backed by evidence."
     },
-    "relevant_file_count": 669,
+    "relevant_file_count": 671,
     "unused_resource_dirs": [],
     "quality_signal_points": 130,
     "quality_density": 131.3
diff --git a/reports/context_budget.md b/reports/context_budget.md
index afd7ad1..7d56a3a 100644
--- a/reports/context_budget.md
+++ b/reports/context_budget.md
@@ -2,7 +2,7 @@
 
 | Target | Path | Tier | Limit | Initial | SKILL | Deferred | Resource Governance | Large Deferred Dirs | Quality Density | Unused Dirs | Status |
 | --- | --- | --- | ---: | ---: | ---: | ---: | --- | --- | ---: | --- | --- |
-| root | `.` | `production` | 1000 | 990 | 797 | 493019 | `governed` | scripts:432921 | 131.3 | - | ok |
+| root | `.` | `production` | 1000 | 990 | 797 | 494370 | `governed` | scripts:434272 | 131.3 | - | ok |
 | complex-release-orchestrator | `examples/complex-release-orchestrator/generated-skill` | `production` | 1000 | 790 | 718 | 1657 | `not-required` | - | 164.6 | - | ok |
 | governed-incident-command | `examples/governed-incident-command/generated-skill` | `production` | 1000 | 760 | 658 | 1030 | `not-required` | - | 171.1 | - | ok |
 
diff --git a/reports/context_budget_summary.json b/reports/context_budget_summary.json
index 6d2fa88..ce753f7 100644
--- a/reports/context_budget_summary.json
+++ b/reports/context_budget_summary.json
@@ -1,5 +1,5 @@
 {
-  "generated_at": "2026-06-16",
+  "generated_at": "2026-03-31",
   "targets": [
     {
       "label": "root",
@@ -8,12 +8,12 @@
       "budget_limit": 1000,
       "initial_tokens": 990,
       "skill_body_tokens": 797,
-      "deferred_resource_tokens": 493019,
+      "deferred_resource_tokens": 494370,
       "large_deferred_resource_dirs": [
         {
           "path": "scripts",
-          "estimated_tokens": 432921,
-          "file_count": 137
+          "estimated_tokens": 434272,
+          "file_count": 138
         }
       ],
       "deferred_resource_governance": {
@@ -35,8 +35,8 @@
             ],
             "missing": [],
             "path": "scripts",
-            "estimated_tokens": 432921,
-            "file_count": 137,
+            "estimated_tokens": 434272,
+            "file_count": 138,
             "rationale": "Script resources are deterministic deferred tools, not initial-load prompt context."
           }
         ],
diff --git a/reports/python_compatibility.json b/reports/python_compatibility.json
index f5e37eb..6515f38 100644
--- a/reports/python_compatibility.json
+++ b/reports/python_compatibility.json
@@ -5,7 +5,7 @@
   "root": ".",
   "summary": {
     "target_python": "3.11",
-    "file_count": 214,
+    "file_count": 215,
     "issue_count": 0,
     "syntax_error_count": 0,
     "fstring_311_violation_count": 0,
@@ -760,6 +760,12 @@
       "issue_count": 0,
       "issues": []
     },
+    {
+      "path": "scripts/world_class_native_permission_evidence.py",
+      "ok": true,
+      "issue_count": 0,
+      "issues": []
+    },
     {
       "path": "scripts/world_class_preflight_layout.py",
       "ok": true,
diff --git a/reports/python_compatibility.md b/reports/python_compatibility.md
index 9d1c8aa..f11e1d7 100644
--- a/reports/python_compatibility.md
+++ b/reports/python_compatibility.md
@@ -6,7 +6,7 @@ Generated at: `2026-06-16`
 
 - decision: `pass`
 - target python: `3.11`
-- files scanned: `214`
+- files scanned: `215`
 - issues: `0`
 - syntax errors: `0`
 - f-string 3.11 violations: `0`
diff --git a/reports/security_trust_report.json b/reports/security_trust_report.json
index c911430..aa724e2 100644
--- a/reports/security_trust_report.json
+++ b/reports/security_trust_report.json
@@ -2,9 +2,9 @@
   "ok": true,
   "skill_dir": ".",
   "summary": {
-    "scanned_files": 228,
-    "script_count": 137,
-    "internal_module_count": 51,
+    "scanned_files": 229,
+    "script_count": 138,
+    "internal_module_count": 52,
     "secret_findings": 0,
     "dependency_files": [
       "requirements-ci.txt"
@@ -22,8 +22,8 @@
     "help_smoke_failed_count": 0,
     "interactive_script_count": 0,
     "package_hash_scope": "source-contract-without-generated-reports",
-    "package_hash_file_count": 228,
-    "package_sha256": "662e1fd1592dfdea9053d3e1e6458c81d71c45a2900dc7814a557fcd8e31f4e0"
+    "package_hash_file_count": 229,
+    "package_sha256": "2bc5e8214814c344534e5155301cd3f86e4ceee351313fb3a538d6fd9d400b75"
   },
   "failures": [],
   "warnings": [],
@@ -1726,6 +1726,20 @@
       "network_urls": [],
       "network_hosts": []
     },
+    {
+      "path": "scripts/world_class_native_permission_evidence.py",
+      "interface": "internal-module",
+      "interface_declared": true,
+      "interface_reason": "Imported by world_class_evidence_contract.py to validate runtime permission evidence from target-level probe rows.",
+      "has_argparse": false,
+      "has_main_guard": false,
+      "uses_input": false,
+      "uses_network": false,
+      "uses_file_write": false,
+      "uses_subprocess": false,
+      "network_urls": [],
+      "network_hosts": []
+    },
     {
       "path": "scripts/world_class_preflight_layout.py",
       "interface": "internal-module",
@@ -1999,7 +2013,7 @@
     "checked_count": 86,
     "passed_count": 86,
     "failed_count": 0,
-    "skipped_count": 51,
+    "skipped_count": 52,
     "failed_scripts": [],
     "results": [
       {
@@ -3004,6 +3018,10 @@
         "path": "scripts/world_class_human_evidence.py",
         "reason": "internal module"
       },
+      {
+        "path": "scripts/world_class_native_permission_evidence.py",
+        "reason": "internal module"
+      },
       {
         "path": "scripts/world_class_preflight_layout.py",
         "reason": "internal module"
diff --git a/reports/security_trust_report.md b/reports/security_trust_report.md
index 1b6c026..557ac00 100644
--- a/reports/security_trust_report.md
+++ b/reports/security_trust_report.md
@@ -1,9 +1,9 @@
 # Security Trust Report
 
 - OK: `True`
-- Scanned files: `228`
-- Scripts: `137`
-- Internal script modules: `51`
+- Scanned files: `229`
+- Scripts: `138`
+- Internal script modules: `52`
 - Secret findings: `0`
 - Network-capable scripts: `3`
 - Network policy covered scripts: `3`
@@ -15,8 +15,8 @@
 - CLI help smoke failures: `0`
 - Interactive scripts: `0`
 - Package hash scope: `source-contract-without-generated-reports`
-- Package hash files: `228`
-- Package SHA256: `662e1fd1592dfdea9053d3e1e6458c81d71c45a2900dc7814a557fcd8e31f4e0`
+- Package hash files: `229`
+- Package SHA256: `2bc5e8214814c344534e5155301cd3f86e4ceee351313fb3a538d6fd9d400b75`
 
 ## Failures
 
@@ -182,6 +182,7 @@
 | scripts/verify_package.py | cli | False | True | True | False | False | True | False | Default CLI classification; add SCRIPT_INTERFACE for internal modules. |
 | scripts/world_class_evidence_contract.py | internal-module | True | False | False | False | False | False | False | Imported by world-class evidence reports to share intake validation and artifact integrity checks. |
 | scripts/world_class_human_evidence.py | internal-module | True | False | False | False | False | False | False | Imported by world_class_evidence_contract.py to validate blind A/B human adjudication evidence from decision rows. |
+| scripts/world_class_native_permission_evidence.py | internal-module | True | False | False | False | False | False | False | Imported by world_class_evidence_contract.py to validate runtime permission evidence from target-level probe rows. |
 | scripts/world_class_preflight_layout.py | internal-module | True | False | False | False | False | False | False | Imported by render_world_class_preflight.py to keep preflight HTML layout out of data assembly. |
 | scripts/world_class_provider_evidence.py | internal-module | True | False | False | False | False | False | False | Imported by world_class_evidence_contract.py to validate provider-backed holdout execution evidence from run rows. |
 | scripts/world_class_source_checks.py | internal-module | True | False | False | False | False | False | False | Imported by world-class evidence reports to keep source-evidence readiness checks consistent. |
diff --git a/reports/skill-os-2-review.md b/reports/skill-os-2-review.md
index 8288708..2b2f91c 100644
--- a/reports/skill-os-2-review.md
+++ b/reports/skill-os-2-review.md
@@ -22,6 +22,7 @@ Yao Meta Skill is no longer only a Meta Skill factory. The current working tree
 - World-Class Evidence Ledger v0 that records current acceptance state, provenance requirements, privacy contracts, and anti-overclaim guards for those remaining gaps.
 - World-Class Intake Contract Hardening v0 so real evidence submissions must use the ledger's canonical `<evidence-key>.json` filename and are recursively rejected when they include raw prompt, output, transcript, message, credential, secret, token, or API-key fields.
 - World-Class Provider Evidence Guard v0 so provider holdout submissions must reconcile `summary` counts with `runs` rows and include at least one passing model run whose provider, model, timing, non-estimated usage, and output hash match the submitted provenance.
+- World-Class Native Permission Evidence Guard v0 so native-permission submissions must reconcile runtime probe summary counts with target rows, require at least one native-enforced target row, and keep installer permission checks failure-free.
 - World-Class Submission Matrix v0 so the submission kit exposes a single operator-facing matrix for draft status, artifact readiness, source-check blockers, and the next action without counting matrix rows as completion evidence.
 - World-Class Submission Artifact Roles v0 so submission kits distinguish `submission-ref` rows that belong in `artifact_refs` from supporting audit assets, reducing operator ambiguity without accepting evidence.
 - World-Class Submission Kit Rendering v0 so Markdown and HTML kit presentation lives in a dedicated internal renderer, keeping the CLI focused on evidence assembly and file emission.
@@ -67,7 +68,7 @@ This is still not the final world-class state. Target-native behavior contracts
 | Skill OS 2.0 Audit | `scripts/render_skill_os2_audit.py`, `reports/skill_os2_audit.md`, `tests/verify_skill_os2_audit.py` | v0 landed |
 | World-Class Evidence Plan | `scripts/render_world_class_evidence_plan.py`, `reports/world_class_evidence_plan.md`, `tests/verify_world_class_evidence_plan.py` | v0 landed |
 | World-Class Evidence Ledger | `scripts/render_world_class_evidence_ledger.py`, `reports/world_class_evidence_ledger.md`, `tests/verify_world_class_evidence_ledger.py` | v0 landed |
-| World-Class Evidence Intake | `scripts/world_class_evidence_contract.py`, `scripts/world_class_provider_evidence.py`, `scripts/render_world_class_evidence_intake.py`, `evidence/world_class/intake.schema.json`, `tests/verify_world_class_evidence_intake.py` with canonical filename, source-artifact validation, provider run-row validation, and nested raw-field rejection | v0 landed |
+| World-Class Evidence Intake | `scripts/world_class_evidence_contract.py`, `scripts/world_class_provider_evidence.py`, `scripts/world_class_native_permission_evidence.py`, `scripts/render_world_class_evidence_intake.py`, `evidence/world_class/intake.schema.json`, `tests/verify_world_class_evidence_intake.py` with canonical filename, source-artifact validation, provider run-row validation, native permission target-row validation, and nested raw-field rejection | v0 landed |
 | World-Class Submission Kit | `scripts/prepare_world_class_submission_kit.py`, `scripts/world_class_submission_matrix.py`, `scripts/world_class_submission_kit_rendering.py`, `tests/verify_world_class_submission_kit.py` with draft, artifact, source-check, next-action matrix evidence, and separated Markdown/HTML rendering | v0 landed |
 | Runtime Conformance | `scripts/run_conformance_suite.py`, `reports/conformance_matrix.md` | v0 landed |
 | Trust & Security | `scripts/trust_check.py`, `reports/security_trust_report.md`, `security/*.md` | v0 landed |
@@ -128,7 +129,7 @@ Next move: add real client or installer permission enforcement integration.
 | Output Eval | `5` cases, with-skill pass rate `100`, baseline pass rate `0`, with file-backed, near-neighbor, boundary coverage, `10` local command-runner execution runs, `0` recorded fixture runs, `0` provider model-executed runs in root release evidence, `10` estimated token counts, provider runner v0 available, `5` blind A/B review pairs, a generated `reports/output_review_decisions.json` template, `0 / 5` reviewer decisions pending, `0` answer keys revealed, and `5` pending answers hidden |
 | Runtime Conformance | `5 / 5` targets passing |
 | Target Compiler | `5 / 5` compiled target contracts generated for OpenAI, Claude, generic, Agent Skills compatible, and VS Code / Copilot outputs, including target permission contracts and target-native behavior contracts |
-| Trust | `0` secret findings, `1` pinned dependency file, `51` declared internal modules, `3 / 3` network-capable scripts covered by bounded host policy, `86 / 86` CLI help smoke checks passing across `137` scripts, source-contract hash scope explicit |
+| Trust | `0` secret findings, `1` pinned dependency file, `52` declared internal modules, `3 / 3` network-capable scripts covered by bounded host policy, `86 / 86` CLI help smoke checks passing across `138` scripts, source-contract hash scope explicit |
 | Permission Governance | `3 / 3` required high-permission capabilities approved, `0` missing, `0` invalid, `0` expired |
 | Runtime Permission Probes | `4 / 4` target adapters probed, `0` native-enforcement adapters, `4` explicit metadata fallbacks, `4` residual risks retained for reviewer visibility |
 | Skill Atlas | `12` scanned skills, `1` actionable root skill, `1` telemetry report, `0` actionable route collisions, `0` actionable owner gaps, `0` actionable stale skills, `0` actionable drift signals, `24` scoped non-actionable issue signals retained for visibility |
diff --git a/reports/skill_os2_audit.json b/reports/skill_os2_audit.json
index b402372..5dde601 100644
--- a/reports/skill_os2_audit.json
+++ b/reports/skill_os2_audit.json
@@ -197,7 +197,7 @@
       "key": "trust-security",
       "label": "Trust Security",
       "status": "pass",
-      "current": "secrets 0; scripts 137; help failures 0",
+      "current": "secrets 0; scripts 138; help failures 0",
       "target": "Secrets, scripts, dependencies, permissions, and package hash are reviewable",
       "evidence": [
         {
diff --git a/reports/skill_os2_audit.md b/reports/skill_os2_audit.md
index bb985cf..60c19dd 100644
--- a/reports/skill_os2_audit.md
+++ b/reports/skill_os2_audit.md
@@ -23,7 +23,7 @@ Generated at: `2026-06-16`
 | Human Adjudication | human-required | 0/5 decisions; pending 5 | Real reviewer decisions recorded before claiming output review completion | Record real A/B choices in the decision template, then regenerate adjudication. |
 | Benchmark Reproducibility | pass | artifacts 25; missing 0; failures 3 | Public methodology, reproducible commands, required artifacts, and failure disclosure are machine-checkable | Keep the manifest current with every benchmark, package, and release evidence change. |
 | Runtime Conformance | pass | 5/5 targets pass | Target package structure, metadata, relative paths, and degradation notes pass | Keep target conformance fixtures updated as platform contracts change. |
-| Trust Security | pass | secrets 0; scripts 137; help failures 0 | Secrets, scripts, dependencies, permissions, and package hash are reviewable | Keep high-permission approvals scoped, expiring, and target-mapped. |
+| Trust Security | pass | secrets 0; scripts 138; help failures 0 | Secrets, scripts, dependencies, permissions, and package hash are reviewable | Keep high-permission approvals scoped, expiring, and target-mapped. |
 | Permission Metadata | pass | 4/4 target probes pass; metadata fallback 4; installer enforcement 4 | Packaged adapters expose explicit permission metadata, residual risks, and installer enforcement evidence when available | Preserve residual-risk notes until real native enforcement exists. |
 | Native Permission Enforcement | external-required | native-enforced targets 0; installer-enforced targets 4 | At least one target/client enforces approved permissions at runtime | Integrate a real target-client or external installer runtime guard before claiming native permission enforcement. |
 | Skill Atlas | pass | 12 skills; actionable collisions 0 | Workspace catalog, route overlap, stale/owner gaps, drift, and no-route opportunities | Feed real drift data into Atlas once client telemetry is installed. |
diff --git a/reports/skill_os2_coverage.json b/reports/skill_os2_coverage.json
index 5968213..8ad273a 100644
--- a/reports/skill_os2_coverage.json
+++ b/reports/skill_os2_coverage.json
@@ -130,7 +130,7 @@
       "label": "Trust Security",
       "status": "pass",
       "objective": "Scripts, dependencies, permissions, secrets, and package hash are reviewable for team distribution.",
-      "current": "137 scripts; secrets 0; help failures 0",
+      "current": "138 scripts; secrets 0; help failures 0",
       "command": "python3 scripts/yao.py trust .",
       "test": "python3 tests/verify_trust_check.py",
       "evidence": [
diff --git a/reports/skill_os2_coverage.md b/reports/skill_os2_coverage.md
index a36caa8..0b97511 100644
--- a/reports/skill_os2_coverage.md
+++ b/reports/skill_os2_coverage.md
@@ -26,7 +26,7 @@ This report maps the Skill OS 2.0 upgrade blueprint to concrete local artifacts,
 | Skill IR | `pass` | schema 2.0.0; targets 5 | `python3 scripts/yao.py skill-ir .` | `python3 tests/verify_skill_ir.py` |
 | Output Eval Lab | `pass` | 5 cases; delta 100.0; execution 10 | `python3 scripts/yao.py output-exec . && python3 scripts/yao.py output-review .` | `python3 tests/verify_output_eval_lab.py` |
 | Runtime Conformance | `pass` | 5/5 targets pass | `python3 scripts/yao.py conformance .` | `python3 tests/verify_conformance_suite.py` |
-| Trust Security | `pass` | 137 scripts; secrets 0; help failures 0 | `python3 scripts/yao.py trust .` | `python3 tests/verify_trust_check.py` |
+| Trust Security | `pass` | 138 scripts; secrets 0; help failures 0 | `python3 scripts/yao.py trust .` | `python3 tests/verify_trust_check.py` |
 | Skill Atlas | `pass` | 12 scanned skills; actionable collisions 0 | `python3 scripts/yao.py skill-atlas --workspace-root .` | `python3 tests/verify_skill_atlas.py` |
 | Registry Distribution | `pass` | archive entries 658; install failures 0 | `python3 scripts/yao.py package . --platform openai --platform claude --platform generic --platform vscode --output-dir dist --zip && python3 scripts/yao.py registry-audit .` | `python3 tests/verify_registry_audit.py` |
 | Review Studio | `pass` | 16 gates; decision review; warnings 3 | `python3 scripts/yao.py review-studio .` | `python3 tests/verify_review_studio.py` |
diff --git a/scripts/world_class_evidence_contract.py b/scripts/world_class_evidence_contract.py
index 87c3934..4fa978a 100644
--- a/scripts/world_class_evidence_contract.py
+++ b/scripts/world_class_evidence_contract.py
@@ -7,6 +7,7 @@ from pathlib import Path
 from typing import Any
 
 from world_class_human_evidence import validate_human_adjudication_report
+from world_class_native_permission_evidence import validate_native_permission_report
 from world_class_provider_evidence import validate_provider_execution_report
 
 
@@ -411,44 +412,7 @@ def validate_native_permission_artifacts(payload: dict[str, Any], errors: list[s
     probes = load_json(paths.get("reports/runtime_permission_probes.json", root / "__missing__"))
     install = load_json(paths.get("reports/install_simulation.json", root / "__missing__"))
     if probes:
-        probe_summary = summary(probes)
-        add_error(errors, probes.get("ok") is True, "native-permission-enforcement runtime probe report ok must be true")
-        add_error(
-            errors,
-            bool(real_int(probe_summary.get("native_enforcement_count")) and probe_summary["native_enforcement_count"] > 0),
-            "native-permission-enforcement runtime probe summary.native_enforcement_count must be >0",
-        )
-        add_error(
-            errors,
-            probe_summary.get("failure_count") == 0,
-            "native-permission-enforcement runtime probe summary.failure_count must be 0",
-        )
-        add_error(
-            errors,
-            probe_summary.get("installer_enforcement_ready") is True,
-            "native-permission-enforcement runtime probe summary.installer_enforcement_ready must be true",
-        )
-    if install:
-        install_summary = summary(install)
-        add_error(errors, install.get("ok") is True, "native-permission-enforcement install simulation report ok must be true")
-        add_error(
-            errors,
-            bool(
-                real_int(install_summary.get("installer_permission_enforced_count"))
-                and install_summary["installer_permission_enforced_count"] > 0
-            ),
-            "native-permission-enforcement install simulation summary.installer_permission_enforced_count must be >0",
-        )
-        add_error(
-            errors,
-            install_summary.get("installer_permission_failure_count") == 0,
-            "native-permission-enforcement install simulation summary.installer_permission_failure_count must be 0",
-        )
-        add_error(
-            errors,
-            install_summary.get("failure_count") == 0,
-            "native-permission-enforcement install simulation summary.failure_count must be 0",
-        )
+        validate_native_permission_report(probes, install, errors)
 
 
 def validate_native_client_telemetry_artifacts(payload: dict[str, Any], errors: list[str], root: Path) -> None:
diff --git a/scripts/world_class_native_permission_evidence.py b/scripts/world_class_native_permission_evidence.py
new file mode 100644
index 0000000..c64de04
--- /dev/null
+++ b/scripts/world_class_native_permission_evidence.py
@@ -0,0 +1,143 @@
+#!/usr/bin/env python3
+from typing import Any
+
+
+SCRIPT_INTERFACE = "internal-module"
+SCRIPT_INTERFACE_REASON = "Imported by world_class_evidence_contract.py to validate runtime permission evidence from target-level probe rows."
+
+
+def add_error(errors: list[str], condition: bool, message: str) -> None:
+    if not condition:
+        errors.append(message)
+
+
+def real_int(value: Any) -> int | None:
+    return value if isinstance(value, int) and not isinstance(value, bool) else None
+
+
+def object_list(value: Any) -> list[dict[str, Any]]:
+    return [item for item in value if isinstance(item, dict)] if isinstance(value, list) else []
+
+
+def sorted_strings(value: Any) -> list[str]:
+    return sorted(str(item) for item in value) if isinstance(value, list) else []
+
+
+def validate_target_rows(probes: dict[str, Any], summary: dict[str, Any], errors: list[str]) -> None:
+    targets = object_list(probes.get("targets", []))
+    target_count = real_int(summary.get("target_count"))
+    native_count = real_int(summary.get("native_enforcement_count"))
+    fail_count = real_int(summary.get("fail_count"))
+    expected_capabilities = sorted_strings(probes.get("expected_capabilities"))
+
+    add_error(errors, bool(targets), "native-permission-enforcement runtime probe targets must not be empty")
+    if target_count is not None:
+        add_error(errors, len(targets) == target_count, "native-permission-enforcement runtime probe targets length must equal summary.target_count")
+    add_error(errors, fail_count == 0, "native-permission-enforcement runtime probe summary.fail_count must be 0")
+
+    observed_native = 0
+    for index, target in enumerate(targets, start=1):
+        target_name = str(target.get("target", "")).strip() or str(index)
+        add_error(errors, target.get("status") == "pass", f"native-permission-enforcement target {target_name} status must be pass")
+        add_error(errors, not target.get("failures"), f"native-permission-enforcement target {target_name} failures must be empty")
+        checks = object_list(target.get("checks", []))
+        add_error(errors, bool(checks), f"native-permission-enforcement target {target_name} checks must not be empty")
+        add_error(
+            errors,
+            all(item.get("passed") is True for item in checks),
+            f"native-permission-enforcement target {target_name} checks must all pass",
+        )
+        native = target.get("native_enforcement") is True
+        if native:
+            observed_native += 1
+            add_error(
+                errors,
+                target.get("assurance") == "native-enforced",
+                f"native-permission-enforcement target {target_name} assurance must be native-enforced",
+            )
+        else:
+            add_error(
+                errors,
+                target.get("metadata_fallback_explicit") is True,
+                f"native-permission-enforcement fallback target {target_name} must keep metadata_fallback_explicit true",
+            )
+            add_error(
+                errors,
+                bool(target.get("residual_risks")),
+                f"native-permission-enforcement fallback target {target_name} must retain residual risks",
+            )
+        if expected_capabilities:
+            add_error(
+                errors,
+                sorted_strings(target.get("declared_capabilities")) == expected_capabilities,
+                f"native-permission-enforcement target {target_name} declared_capabilities must match expected_capabilities",
+            )
+        installer = target.get("installer_enforcement", {}) if isinstance(target.get("installer_enforcement"), dict) else {}
+        add_error(
+            errors,
+            installer.get("enforced") is True,
+            f"native-permission-enforcement target {target_name} installer_enforcement.enforced must be true",
+        )
+        add_error(
+            errors,
+            not installer.get("missing_capabilities"),
+            f"native-permission-enforcement target {target_name} installer_enforcement missing_capabilities must be empty",
+        )
+        add_error(
+            errors,
+            not installer.get("failure_details"),
+            f"native-permission-enforcement target {target_name} installer_enforcement failure_details must be empty",
+        )
+
+    add_error(errors, bool(native_count and observed_native == native_count), "native-permission-enforcement native target rows must match summary.native_enforcement_count")
+    add_error(errors, observed_native > 0, "native-permission-enforcement must include at least one native-enforced target row")
+
+
+def validate_install_simulation_report(install: dict[str, Any], errors: list[str]) -> None:
+    summary = install.get("summary", {}) if isinstance(install.get("summary", {}), dict) else {}
+    add_error(errors, install.get("ok") is True, "native-permission-enforcement install simulation report ok must be true")
+    add_error(
+        errors,
+        bool(real_int(summary.get("installer_permission_enforced_count")) and summary["installer_permission_enforced_count"] > 0),
+        "native-permission-enforcement install simulation summary.installer_permission_enforced_count must be >0",
+    )
+    add_error(
+        errors,
+        summary.get("installer_permission_failure_count") == 0,
+        "native-permission-enforcement install simulation summary.installer_permission_failure_count must be 0",
+    )
+    add_error(errors, summary.get("failure_count") == 0, "native-permission-enforcement install simulation summary.failure_count must be 0")
+    permission_checks = [
+        item
+        for item in object_list(install.get("checks", []))
+        if str(item.get("id", "")).startswith("permission-") and str(item.get("id", "")).endswith(("-approved", "-target-enforcement"))
+    ]
+    if permission_checks:
+        add_error(
+            errors,
+            all(item.get("status") == "pass" for item in permission_checks),
+            "native-permission-enforcement install simulation permission checks must all pass",
+        )
+
+
+def validate_native_permission_report(probes: dict[str, Any], install: dict[str, Any], errors: list[str]) -> None:
+    summary = probes.get("summary", {}) if isinstance(probes.get("summary", {}), dict) else {}
+    add_error(errors, probes.get("ok") is True, "native-permission-enforcement runtime probe report ok must be true")
+    add_error(
+        errors,
+        bool(real_int(summary.get("native_enforcement_count")) and summary["native_enforcement_count"] > 0),
+        "native-permission-enforcement runtime probe summary.native_enforcement_count must be >0",
+    )
+    add_error(
+        errors,
+        summary.get("failure_count") == 0,
+        "native-permission-enforcement runtime probe summary.failure_count must be 0",
+    )
+    add_error(
+        errors,
+        summary.get("installer_enforcement_ready") is True,
+        "native-permission-enforcement runtime probe summary.installer_enforcement_ready must be true",
+    )
+    validate_target_rows(probes, summary, errors)
+    if install:
+        validate_install_simulation_report(install, errors)
diff --git a/tests/verify_world_class_evidence_intake.py b/tests/verify_world_class_evidence_intake.py
index e378f9b..7c9794d 100644
--- a/tests/verify_world_class_evidence_intake.py
+++ b/tests/verify_world_class_evidence_intake.py
@@ -201,16 +201,47 @@ def provider_artifact_submission(skill_root: Path) -> dict:
 
 
 def write_native_permission_artifacts(skill_root: Path, *, complete: bool) -> None:
+    capabilities = ["file_write", "network", "subprocess"]
+    target = {
+        "target": "vscode",
+        "status": "pass",
+        "adapter": "dist/targets/vscode/adapter.json",
+        "permission_model": "native-client",
+        "native_enforcement": complete,
+        "metadata_fallback_explicit": not complete,
+        "installer_enforcement": {
+            "target": "vscode",
+            "source_status": "present",
+            "enforced": True,
+            "enforced_capabilities": capabilities,
+            "missing_capabilities": [],
+            "failure_details": [],
+        },
+        "assurance": "native-enforced" if complete else "metadata-fallback-explicit",
+        "declared_capabilities": capabilities,
+        "checks": [
+            {"key": "adapter-present", "passed": True, "detail": "vscode adapter.json is readable"},
+            {"key": "native-enforcement-boolean", "passed": True, "detail": "native flag is boolean"},
+        ],
+        "failures": [],
+        "residual_risks": [] if complete else ["Client-native permission enforcement is not provided by this target."],
+    }
     write_json(
         skill_root / "reports" / "runtime_permission_probes.json",
         {
             "schema_version": "1.0",
             "ok": True,
+            "expected_capabilities": capabilities,
             "summary": {
+                "target_count": 1,
+                "pass_count": 1,
+                "fail_count": 0,
                 "native_enforcement_count": 1 if complete else 0,
+                "metadata_fallback_count": 0 if complete else 1,
                 "failure_count": 0,
                 "installer_enforcement_ready": True,
             },
+            "targets": [target],
         },
     )
     write_json(
@@ -221,8 +252,26 @@ def write_native_permission_artifacts(skill_root: Path, *, complete: bool) -> No
             "summary": {
                 "installer_permission_enforced_count": 12,
                 "installer_permission_failure_count": 0,
+                "permission_target_count": 1,
+                "permission_capability_count": 3,
                 "failure_count": 0,
             },
+            "checks": [
+                {"id": "permission-vscode-file_write-approved", "status": "pass", "detail": "file write approved"},
+                {
+                    "id": "permission-vscode-file_write-target-enforcement",
+                    "status": "pass",
+                    "detail": "file write enforced",
+                },
+                {"id": "permission-vscode-network-approved", "status": "pass", "detail": "network approved"},
+                {"id": "permission-vscode-network-target-enforcement", "status": "pass", "detail": "network enforced"},
+                {"id": "permission-vscode-subprocess-approved", "status": "pass", "detail": "subprocess approved"},
+                {
+                    "id": "permission-vscode-subprocess-target-enforcement",
+                    "status": "pass",
+                    "detail": "subprocess enforced",
+                },
+            ],
         },
     )
 
@@ -415,6 +464,23 @@ def assert_external_contract_artifact_validation() -> None:
         template_expected=False,
     )
     assert permission_valid["status"] == "pass", permission_valid
+    forged_probe = json.loads((skill_root / "reports" / "runtime_permission_probes.json").read_text(encoding="utf-8"))
+    forged_probe["targets"][0]["native_enforcement"] = False
+    forged_probe["targets"][0]["assurance"] = "metadata-fallback-explicit"
+    forged_probe["targets"][0]["metadata_fallback_explicit"] = True
+    forged_probe["targets"][0]["residual_risks"] = ["Fallback still requires operator enforcement."]
+    write_json(skill_root / "reports" / "runtime_permission_probes.json", forged_probe)
+    permission_forged = validate_payload(
+        native_permission_submission(skill_root),
+        permission_entry,
+        path=skill_root / "evidence" / "world_class" / "submissions" / "native-permission-enforcement.json",
+        root=skill_root,
+        template_expected=False,
+    )
+    assert permission_forged["status"] == "fail", permission_forged
+    assert any("native target rows must match summary.native_enforcement_count" in error for error in permission_forged["errors"]), (
+        permission_forged["errors"]
+    )
     write_native_permission_artifacts(skill_root, complete=False)
     permission_invalid = validate_payload(
         native_permission_submission(skill_root),