gwonxhj · hyeokjun32 · May 29, 2026 · May 29, 2026
diff --git a/README.ko.md b/README.ko.md
@@ -213,6 +213,7 @@ Runtime Intelligence smoke:
 - Orchestrator operation feed를 supplemental context로 보존
 - EdgeEnv telemetry history/regression evidence를 Lab report에 연결
 - Runtime history seed `run_config` snapshot을 replay/comparability traceability로 표시
+- Jetson EdgeEnv preservation smoke는 entrypoint가 device-local ONNX Runtime probe evidence, live `tegrastats`, Runtime operation summary를 EdgeEnv run evidence로 보존한 뒤 Lab deployment risk report까지 연결할 수 있음을 확인합니다.
 - 사용 가능한 경우 AIGuard deterministic runtime evidence 보존
 - AIGuard `runtime_history_seed_run_config_traceability` evidence를 gate에서 필수 traceability evidence로 검증
 - AIGuard raw context의 device-local producer lineage를 traceability evidence로 표시
@@ -229,6 +230,25 @@ Runtime Intelligence report에서 읽어야 할 핵심 row:
 - AIGuard deterministic anomaly evidence
 - Lab-owned deployment decision
 
+최신 Jetson EdgeEnv preservation smoke:
+
+| Evidence | Value |
+|---|---:|
+| Operation path | `device_local_starter` |
+| Frames | 32 |
+| Max queue depth | 6 |
+| Dropped / fallback count | 29 / 29 |
+| Deadline missed count | 18 |
+| Parsed `tegrastats` samples | 4 |
+| Max temperature / RAM | 42.843 C / 999 MB |
+| Vision mean / p95 latency | 166.941 ms / 423.192 ms |
+| EdgeEnv run ID | `run-20260529-034704-fbf753f0` |
+| EdgeEnv summary | `runtime_operation_summary` stored |
+| AIGuard verdict | `blocked` / `high` |
+| Lab decision | `blocked` |
+
+이 기록은 Runtime Intelligence handoff를 위한 device-local starter smoke입니다. EdgeEnv는 local run evidence와 supplemental operation context를 보존하고, Lab은 final deployment decision owner로 남습니다. decoded YOLO accuracy validation, live camera operation, production remote execution, thermal endurance validation으로 해석하지 않습니다.
+
 재현 smoke:
 
 ```bash

diff --git a/README.md b/README.md
@@ -112,6 +112,7 @@ Runtime Intelligence smoke:
 - Orchestrator operation feed preserved as supplemental context
 - EdgeEnv telemetry history/regression evidence ingested by Lab
 - EdgeEnv handoff gate checks device-local producer lineage in preserved Orchestrator context
+- Jetson EdgeEnv preservation smoke confirms the entrypoint can carry device-local ONNX Runtime probe evidence, live `tegrastats`, and Runtime operation summary into EdgeEnv run evidence before Lab renders the deployment risk report
 - AIGuard deterministic runtime evidence preserved when available
 - Lab-owned Runtime Intelligence Risk Summary generated without changing existing JSON contracts
 
@@ -518,6 +519,29 @@ Markdown and HTML reports include a Runtime Intelligence Risk Summary that conne
 - AIGuard deterministic evidence
 - the Lab-owned deployment decision
 
+Latest Jetson EdgeEnv preservation smoke:
+
+| Evidence | Value |
+|---|---:|
+| Operation path | `device_local_starter` |
+| Frames | 32 |
+| Max queue depth | 6 |
+| Dropped / fallback count | 29 / 29 |
+| Deadline missed count | 18 |
+| Parsed `tegrastats` samples | 4 |
+| Max temperature / RAM | 42.843 C / 999 MB |
+| Vision mean / p95 latency | 166.941 ms / 423.192 ms |
+| EdgeEnv run ID | `run-20260529-034704-fbf753f0` |
+| EdgeEnv summary | `runtime_operation_summary` stored |
+| AIGuard verdict | `blocked` / `high` |
+| Lab decision | `blocked` |
+
+This is device-local starter smoke evidence for the Runtime Intelligence
+handoff. EdgeEnv preserves local run evidence and supplemental operation
+context; Lab remains the final deployment decision owner. It is not decoded
+YOLO accuracy validation, live camera operation, production remote execution, or
+thermal endurance validation.
+
 This does not change existing JSON contracts. Lab fixtures for these cases live under `tests/fixtures/edgeenv_regression/`.
 
 #### Local Smoke

diff --git a/docs/portfolio/edgeenv_runtime_regression_lab_handoff.md b/docs/portfolio/edgeenv_runtime_regression_lab_handoff.md
@@ -38,6 +38,44 @@ Expected ownership:
 
 This makes the portfolio chain stronger without turning Lab into a production observability dashboard or making Orchestrator/AIGuard final decision owners.
 
+## Jetson EdgeEnv preservation smoke
+
+The entrypoint path was replayed on Jetson with `--device-local`,
+`--capture-tegrastats`, and `--edgeenv-run-evidence` so the Lab-facing Runtime
+Intelligence chain can be traced from real device-local operation evidence into
+EdgeEnv's local run registry before the Lab deployment risk report is rendered.
+
+```text
+InferEdge entrypoint device-local run
+-> Orchestrator operation evidence
+-> EdgeEnv local run evidence registry
+-> AIGuard deterministic runtime warning evidence
+-> Lab-owned agent runtime deployment risk report
+```
+
+Observed Jetson smoke values:
+
+| Evidence | Value |
+|---|---:|
+| Operation path | `device_local_starter` |
+| Frames | 32 |
+| Max queue depth | 6 |
+| Dropped / fallback count | 29 / 29 |
+| Deadline missed count | 18 |
+| Parsed `tegrastats` samples | 4 |
+| Max temperature / RAM | 42.843 C / 999 MB |
+| Vision mean / p95 latency | 166.941 ms / 423.192 ms |
+| EdgeEnv run ID | `run-20260529-034704-fbf753f0` |
+| EdgeEnv summary | `runtime_operation_summary` stored |
+| AIGuard verdict | `blocked` / `high` |
+| Lab decision | `blocked` |
+
+This is device-local starter smoke evidence. EdgeEnv remains the local-first
+run evidence registry and supplemental operation-context preservation layer.
+Lab remains the final deployment decision owner. This record is not decoded
+YOLO accuracy validation, live camera operation, production remote execution,
+or thermal endurance validation.
+
 ## Fixture
 
 Committed lightweight fixtures:

diff --git a/docs/portfolio/inferedge_portfolio_submission.md b/docs/portfolio/inferedge_portfolio_submission.md
@@ -144,6 +144,7 @@ Recent validation evidence:
 - YOLOv8 COCO subset evaluation: a 10-image local person-detection subset with 89 ground-truth boxes is converted into a COCO-style annotation fixture and evaluated through the `yolov8_coco` preset. The generated report records metric backend `simplified`, mAP@50 0.1410, precision 0.2941, recall 0.1685, and structural validation passed. This is documented as subset workflow evidence, not a full COCO benchmark claim. `pycocotools` remains an optional explicit backend.
 - Validation problem cases: the demo bundle includes annotation-missing, invalid detection structure, contract shape mismatch, and latency regression reports. These show that InferEdge records review/block evidence explicitly instead of presenting every validation path as successful.
 - Runtime Intelligence smoke chain: the committed bundle manifest and report gates verify Orchestrator `edgeenv_runtime_telemetry_feed`, EdgeEnv `runtime_telemetry_context.history.telemetry_coverage`, AIGuard deterministic runtime operation evidence, and Lab Runtime Intelligence Risk Summary ownership. This is a local-first artifact chain, not production observability or a runtime control plane.
+- Jetson EdgeEnv preservation smoke: the InferEdge entrypoint replayed a 32-frame `device_local_starter` path on Jetson with a user-provided `yolov8n.onnx`, live `tegrastats`, process resource snapshot capture, and `--edgeenv-run-evidence`. The run reached max queue depth `6`, recorded dropped/fallback `29 / 29`, deadline misses `18`, parsed `4` `tegrastats` samples, observed max temperature / RAM `42.843 C / 999 MB`, stored EdgeEnv run `run-20260529-034704-fbf753f0` with `runtime_operation_summary`, and produced AIGuard `blocked/high` plus Lab `blocked` deployment risk evidence. This is device-local starter smoke, not decoded YOLO accuracy, live camera, production remote execution, or thermal endurance validation.
 
 The direct Runtime execution result includes `deployment_decision`. Its `unknown` value is expected before Lab compare/report because the worker response has not yet been compared by Lab.
 
@@ -170,6 +171,7 @@ Forge summary
 - **SaaS-ready API + async job workflow:** Lab has API response contracts, in-memory async job stubs, and worker request/response mapping without prematurely adding DB/queue infrastructure.
 - **Deterministic rule-based diagnosis:** AIGuard uses rule + evidence detectors instead of vague LLM judgement.
 - **Runtime Intelligence evidence chain:** Orchestrator operation context, EdgeEnv telemetry coverage/regression, AIGuard deterministic anomaly evidence, and Lab deployment risk reporting are connected through lightweight fixtures and gates without adding a new repo or production monitoring platform.
+- **Jetson Runtime Intelligence preservation evidence:** device-local ONNX Runtime probe and live `tegrastats` evidence can be preserved through EdgeEnv's local run registry and rendered back into a Lab-owned deployment risk report without changing existing Runtime result or compare contracts.
 - **Deployment decision ownership:** Lab keeps final deploy/review/blocked ownership while preserving optional guard evidence.
 - **Local-first Studio demo:** The browser UI can replay real validation evidence locally without adding DB, queue, upload, auth, billing, or production SaaS infrastructure.
 

diff --git a/tests/test_runtime_intelligence_bundle_manifest.py b/tests/test_runtime_intelligence_bundle_manifest.py
@@ -140,6 +140,27 @@ def test_runtime_intelligence_docs_describe_expected_report_markers():
             assert marker in doc
 
 
+def test_runtime_intelligence_docs_record_jetson_edgeenv_preservation_boundary():
+    docs = [
+        REPO_ROOT / "README.md",
+        REPO_ROOT / "README.ko.md",
+        REPO_ROOT / "docs" / "portfolio" / "edgeenv_runtime_regression_lab_handoff.md",
+        REPO_ROOT / "docs" / "portfolio" / "inferedge_portfolio_submission.md",
+    ]
+
+    for path in docs:
+        text = path.read_text(encoding="utf-8")
+        assert "Jetson EdgeEnv preservation smoke" in text
+        assert "device_local_starter" in text
+        assert "run-20260529-034704-fbf753f0" in text
+        assert "runtime_operation_summary" in text
+        assert "29 / 29" in text
+        assert "42.843 C / 999 MB" in text
+        assert "Lab" in text and "decision owner" in text
+        assert "production remote execution" in text
+        assert "thermal endurance validation" in text
+
+
 def test_runtime_intelligence_bundle_manifest_gate_validates_edgeenv_handoff(
     tmp_path,
 ):