docs: make runtime intelligence README scannable

README.ko.md

@@ -220,30 +220,33 @@ Cross-repo evidence:
 
 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
-- 기존 JSON contract를 바꾸지 않고 Lab-owned Runtime Intelligence Risk Summary 생성
-
-EdgeEnv runtime regression report에 `runtime_telemetry_context`가 포함되면 Lab은 이를 supplemental telemetry coverage / evidence-gap context로 표시하되, final deployment decision ownership은 Lab에 남깁니다.
-
-Runtime Intelligence report에서 읽어야 할 핵심 row:
-
-- EdgeEnv comparability / regression evidence
-- telemetry replay gap
-- Runtime history seed `run_config` traceability
-- Orchestrator device-local producer lineage
-- Lab EdgeEnv preservation context marker
-- Jetson/device-local preservation row의 `identity=jetson_device_local_preservation`
-  및 `path=device_local_starter` 식별 label
-- Jetson/device-local preservation detail row의 producer/source/stage/resource
-  navigation context
-- AIGuard deterministic anomaly evidence
-- Lab-owned deployment decision
+- Orchestrator operation feed를 supplemental context로 보존합니다.
+- EdgeEnv telemetry history/regression evidence를 Lab report에 연결합니다.
+- AIGuard deterministic runtime evidence가 있으면 같은 Lab-owned report에 보존합니다.
+- Jetson EdgeEnv preservation smoke는 device-local ONNX Runtime probe evidence, live `tegrastats`, Runtime operation summary가 EdgeEnv run evidence를 거쳐 Lab deployment risk report까지 이어지는지 확인합니다.
+- 기존 JSON contract를 바꾸지 않고 Lab-owned Runtime Intelligence Risk Summary를 생성합니다.
+
+### Runtime Intelligence Risk Summary 빠른 읽기
+
+| 먼저 볼 항목 | Quick signal | 의미 |
+|---|---|---|
+| Decision owner | `Lab remains the final deployment decision owner` | EdgeEnv, AIGuard, Orchestrator는 evidence provider이고 최종 판단은 Lab이 소유합니다. |
+| EdgeEnv regression gate | EdgeEnv comparability / regression evidence | runtime regression은 EdgeEnv comparability context가 있을 때만 해석합니다. |
+| Telemetry/replay quality | telemetry replay gap, `runtime_history_seed_run_config_traceability` | Runtime history seed와 `run_config` traceability가 보존됐는지 확인합니다. |
+| Operation context | `Orchestrator queue/deadline/fallback markers` | queue pressure, `max_total_queue_depth`, deadline miss, fallback count를 한눈에 묶습니다. |
+| AIGuard warnings | deterministic AIGuard runtime operation evidence | AIGuard warning은 Lab policy를 덮어쓰지 않는 review evidence입니다. |
+
+Marker group:
+
+| 그룹 | 핵심 row / label | 이유 |
+|---|---|---|
+| Producer lineage | `edgeenv_orchestrator_producer_lineage`, `runtime_history_seed_run_config_traceability` | EdgeEnv/Orchestrator lineage가 AIGuard와 Lab까지 보존됐는지 확인합니다. |
+| Queue pressure | `Orchestrator queue/deadline/fallback markers`, `AIGuard max queue raw-context traceability` | `max_total_queue_depth`가 AIGuard deterministic raw context와 연결되는지 보여줍니다. |
+| Replay / preservation | `Runtime replay duration scope`, `Lab EdgeEnv preservation context`, `Jetson/device-local EdgeEnv preservation run`, `Jetson/device-local EdgeEnv preservation details` | replay duration과 `identity=jetson_device_local_preservation`, `path=device_local_starter` label을 빠르게 찾게 합니다. |
+| Task / operation risk | `Orchestrator task event rollup`, `AIGuard task event rollup evidence`, `AIGuard runtime operation anomalies` | scheduler delay, deadline miss, fallback decision, queue/drop reason을 review context로 보여줍니다. |
+| Remote starter boundary | `AIGuard remote dispatch event summary`, `Remote fallback starter evidence`, `production_remote_execution=false` | remote dispatch를 production execution이 아니라 starter evidence로 제한합니다. |
+
+세부 marker contract는 [docs/portfolio/edgeenv_runtime_regression_lab_handoff.md](docs/portfolio/edgeenv_runtime_regression_lab_handoff.md)에 정리되어 있습니다.
 
 최신 Jetson EdgeEnv preservation smoke:
 

README.md

@@ -515,46 +515,38 @@ If the EdgeEnv report includes `runtime_telemetry_context`, Lab shows supplement
 
 When `--with-guard` is used with EdgeEnv evidence, Lab preserves deterministic
 AIGuard evidence in the same Lab-owned report. The summary is a reviewer
-navigation surface: it makes runtime risk evidence easy to find, while Lab
-remains the final deployment decision owner.
+navigation surface, not a new owner: Lab remains the final deployment decision
+owner.
 
-The report starts with a `Reviewer Focus` table using
-`Focus / Quick signal / First read` columns so reviewers can quickly scan the
-Decision owner, EdgeEnv regression gate, Telemetry/replay quality,
-Operation context, and AIGuard warnings before reading the detailed evidence
-rows.
+Start with the report's `Reviewer Focus` table. It uses
+`Focus / Quick signal / First read` columns so reviewers can scan the main risk
+areas before opening detailed evidence rows.
 
-| Reviewer question | Where to look | Meaning |
+| First read | Quick signal | What it tells you |
 |---|---|---|
-| Did AIGuard preserve EdgeEnv/Orchestrator producer lineage? | `producer_lineage_evidence_type=edgeenv_orchestrator_producer_lineage` and `runtime_history_seed_run_config_traceability` | Confirms the expected deterministic AIGuard evidence reached the Lab-owned report. |
-| Is there operation pressure? | `Orchestrator queue/deadline/fallback markers` | Keeps `queue_pressure_reason`, `max_total_queue_depth`, deadline misses, and fallback count together before the detailed task rollup. |
-| Can queue depth be traced back to AIGuard raw context? | `AIGuard max queue raw-context traceability` | Ties visible `max_total_queue_depth` markers back to deterministic AIGuard context and Orchestrator operation evidence. |
-| What replay window was used? | `Runtime replay duration scope` | Shows `duration_label`, `duration_class`, frame count, `duration_source`, and `duration_scope_label` such as `source=entrypoint_requested_frames`. |
-| Was EdgeEnv preservation rendered by Lab? | `Lab EdgeEnv preservation context` | Shows `lab_report_preservation_context_present=True`, `lab_preservation=present`, and `lab_context=present` without making EdgeEnv the decision owner. |
-| Is this the Jetson/device-local preservation path? | `Jetson/device-local EdgeEnv preservation run` | Starts with `identity=jetson_device_local_preservation` and `path=device_local_starter` when available. |
-| Where are producer/source/resource details? | `Jetson/device-local EdgeEnv preservation details` | Keeps source, stage, device-local event, resource, and queue markers out of the short identity row. |
-| Which tasks were affected? | `Orchestrator task event rollup` and `AIGuard task event rollup evidence` | Shows scheduler delay, deadline misses, fallback decisions, and queue/drop reasons as review context only. |
-| Is there deterministic operation-risk evidence? | `AIGuard runtime operation anomalies` and `AIGuard Orchestrator Operation Evidence` | Preserves warning evidence without overriding EdgeEnv comparability or Lab deployment policy. |
-| Is this remote dispatch starter evidence? | `AIGuard remote dispatch event summary` and `Remote fallback starter evidence` | Preserves `evidence_role=remote_dispatch_runtime_event_compact_summary`, `operation_boundary=remote dispatch starter evidence only`, and `production_remote_execution=false`. |
+| Decision owner | `Lab remains the final deployment decision owner` | The report is Lab-owned; EdgeEnv, AIGuard, and Orchestrator provide evidence only. |
+| EdgeEnv regression gate | EdgeEnv comparability / regression evidence | Runtime regression is interpreted only after EdgeEnv comparability context is present. |
+| Telemetry/replay quality | telemetry replay gaps and `runtime_history_seed_run_config_traceability` | Shows whether Runtime history seed and `run_config` traceability are available. |
+| Operation context | `Orchestrator queue/deadline/fallback markers` | Groups queue pressure, `max_total_queue_depth`, deadline misses, and fallback count. |
+| AIGuard warnings | deterministic AIGuard runtime operation evidence | Preserves warning evidence without overriding Lab deployment policy. |
+
+Reviewer marker map:
+
+| Marker group | Key rows / labels | Why it matters |
+|---|---|---|
+| Producer lineage | `edgeenv_orchestrator_producer_lineage`, `runtime_history_seed_run_config_traceability` | Confirms EdgeEnv/Orchestrator lineage reached AIGuard and Lab. |
+| Queue pressure | `Orchestrator queue/deadline/fallback markers`, `AIGuard max queue raw-context traceability` | Makes `max_total_queue_depth` traceable back to deterministic AIGuard raw context. |
+| Replay and preservation | `Runtime replay duration scope`, `Lab EdgeEnv preservation context`, `Jetson/device-local EdgeEnv preservation run`, `Jetson/device-local EdgeEnv preservation details` | Shows replay duration, Lab preservation, and `identity=jetson_device_local_preservation` / `path=device_local_starter` labels. |
+| Task and operation risk | `Orchestrator task event rollup`, `AIGuard task event rollup evidence`, `AIGuard runtime operation anomalies` | Shows scheduler delay, deadline misses, fallback decisions, and queue/drop reasons as review context only. |
+| Remote starter boundary | `AIGuard remote dispatch event summary`, `Remote fallback starter evidence`, `production_remote_execution=false` | Keeps remote dispatch as starter evidence, not production remote execution. |
 
 These rows do not turn Orchestrator, EdgeEnv, or AIGuard into final decision
 owners. They also do not present the path as production remote execution,
 long-lived worker readiness, cloud orchestration, or production readiness.
 
-Markdown and HTML reports include a Runtime Intelligence Risk Summary that connects:
-
-- EdgeEnv comparability and regression evidence
-- telemetry replay gaps
-- Runtime history seed and run_config traceability
-- Orchestrator operation risk summary markers
-- compact queue/deadline/fallback operation markers with `max_total_queue_depth`
-- AIGuard raw-context traceability for `max_total_queue_depth`
-- Lab EdgeEnv preservation context markers
-- device-local producer lineage handoff
-- Orchestrator-declared downstream guard alignment
-- remote dispatch starter boundary evidence
-- AIGuard deterministic evidence
-- the Lab-owned deployment decision
+For the full marker contract, see
+[docs/portfolio/edgeenv_runtime_regression_lab_handoff.md](docs/portfolio/edgeenv_runtime_regression_lab_handoff.md)
+([한국어: EdgeEnv 런타임 회귀 Lab handoff 문서](docs/portfolio/edgeenv_runtime_regression_lab_handoff.md)).
 
 Latest Jetson EdgeEnv preservation smoke:
 

tests/test_runtime_intelligence_bundle_manifest.py

@@ -172,7 +172,8 @@ def test_runtime_intelligence_docs_record_jetson_edgeenv_preservation_boundary()
 def test_readme_runtime_intelligence_section_stays_scannable():
     readme = (REPO_ROOT / "README.md").read_text(encoding="utf-8")
 
-    assert "| Reviewer question | Where to look | Meaning |" in readme
+    assert "| First read | Quick signal | What it tells you |" in readme
+    assert "| Marker group | Key rows / labels | Why it matters |" in readme
     assert "Reviewer Focus" in readme
     assert "`Focus / Quick signal / First read`" in readme
     assert "Decision owner" in readme
@@ -181,16 +182,11 @@ def test_readme_runtime_intelligence_section_stays_scannable():
     assert "Operation context" in readme
     assert "AIGuard warnings" in readme
     for row in [
-        "Did AIGuard preserve EdgeEnv/Orchestrator producer lineage?",
-        "Is there operation pressure?",
-        "Can queue depth be traced back to AIGuard raw context?",
-        "What replay window was used?",
-        "Was EdgeEnv preservation rendered by Lab?",
-        "Is this the Jetson/device-local preservation path?",
-        "Where are producer/source/resource details?",
-        "Which tasks were affected?",
-        "Is there deterministic operation-risk evidence?",
-        "Is this remote dispatch starter evidence?",
+        "Producer lineage",
+        "Queue pressure",
+        "Replay and preservation",
+        "Task and operation risk",
+        "Remote starter boundary",
     ]:
         assert row in readme