NVIDIA · mosheabr · Aug 17, 2026 · Aug 17, 2026
@@ -132,7 +132,7 @@ For non-interactive installs, global installs, agent-specific installs, updates,
 | **Holoscan Sensor Bridge** | Agent-ready skills for Holoscan Sensor Bridge devkit workflows, including demo environment bring-up, FPGA flashing for Lattice and VB1940 hardware, example application execution, QA test-plan automation, and support for configuring and using the Holoscan Sensor Bridge FPGA intellectual property (IP) core. | [`hsb-setup`](skills/hsb-setup), [`hsb-flash`](skills/hsb-flash), [`hsb-app`](skills/hsb-app), [`hsb-test`](skills/hsb-test), [`hsb-ip-def`](skills/hsb-ip-def), [`hsb-ip-packetizer`](skills/hsb-ip-packetizer), [`hsb-ip-create-top`](skills/hsb-ip-create-top) |
 | **Isaac for Healthcare Workflows** | Agent-ready skills for Isaac for Healthcare agentic and catheter-navigation workflows, covering task authoring, data pipelines, policy training and validation, CT-derived digital twins, DRR rendering, and interactive catheter simulation. | [`i4h-workflow`](skills/i4h-workflow), [`i4h-workflow-setup`](skills/i4h-workflow-setup), [`i4h-workflow-create`](skills/i4h-workflow-create), [`i4h-workflow-scene-edit`](skills/i4h-workflow-scene-edit), [`i4h-workflow-dataset-teleop`](skills/i4h-workflow-dataset-teleop), [`i4h-workflow-dataset-replay`](skills/i4h-workflow-dataset-replay), [`i4h-workflow-dataset-mimic`](skills/i4h-workflow-dataset-mimic), [`i4h-workflow-dataset-annotate`](skills/i4h-workflow-dataset-annotate), [`i4h-workflow-dataset-convert`](skills/i4h-workflow-dataset-convert), [`i4h-workflow-finetune`](skills/i4h-workflow-finetune), [`i4h-workflow-validate`](skills/i4h-workflow-validate), [`i4h-workflow-e2e`](skills/i4h-workflow-e2e), [`i4h-lerobot-viz`](skills/i4h-lerobot-viz), [`i4h-catheter-navigation`](skills/i4h-catheter-navigation), [`i4h-catheter-navigation-setup`](skills/i4h-catheter-navigation-setup), [`i4h-catheter-navigation-digital-twin`](skills/i4h-catheter-navigation-digital-twin), [`i4h-catheter-navigation-render-drr`](skills/i4h-catheter-navigation-render-drr), [`i4h-catheter-navigation-viewport`](skills/i4h-catheter-navigation-viewport), [`i4h-catheter-navigation-smoke`](skills/i4h-catheter-navigation-smoke), [`i4h-catheter-navigation-e2e`](skills/i4h-catheter-navigation-e2e) |
 | **Jetson BSP** | Agentic skills for setting up and customizing an NVIDIA Jetson Linux Board Support Package (BSP) — pick a target, prepare image and sources, customize IO (camera, PCIe, USB, pinmux, clocks, and more), then promote, flash, and validate. | [`jetson-build-source`](skills/jetson-build-source), [`jetson-customize-camera`](skills/jetson-customize-camera), [`jetson-customize-clocks`](skills/jetson-customize-clocks), [`jetson-customize-fan`](skills/jetson-customize-fan), [`jetson-customize-mgbe`](skills/jetson-customize-mgbe), [`jetson-customize-nvpmodel`](skills/jetson-customize-nvpmodel), [`jetson-customize-pcie`](skills/jetson-customize-pcie), [`jetson-customize-pinmux`](skills/jetson-customize-pinmux), [`jetson-customize-uphy`](skills/jetson-customize-uphy), [`jetson-customize-usb`](skills/jetson-customize-usb), [`jetson-derive-carrier`](skills/jetson-derive-carrier), [`jetson-download-bsp`](skills/jetson-download-bsp), [`jetson-flash-image`](skills/jetson-flash-image), [`jetson-generate-kb`](skills/jetson-generate-kb), [`jetson-init-image`](skills/jetson-init-image), [`jetson-init-source`](skills/jetson-init-source), [`jetson-init-target`](skills/jetson-init-target), [`jetson-link-docs`](skills/jetson-link-docs), [`jetson-optimize-memory`](skills/jetson-optimize-memory), [`jetson-print-bsp-info`](skills/jetson-print-bsp-info), [`jetson-promote-image`](skills/jetson-promote-image), [`jetson-quick-start`](skills/jetson-quick-start), [`jetson-set-target`](skills/jetson-set-target), [`jetson-validate-image`](skills/jetson-validate-image) |
-| **Jetson Device** | Device-side agent skills for working with a live NVIDIA Jetson after boot — diagnostics, memory auditing, headless setup, inference memory tuning, LLM serving and benchmarking, packaging guidance, and speculative decoding. | [`jetson-diagnostic`](skills/jetson-diagnostic), [`jetson-headless-mode`](skills/jetson-headless-mode), [`jetson-inference-mem-tune`](skills/jetson-inference-mem-tune), [`jetson-llm-benchmark`](skills/jetson-llm-benchmark), [`jetson-llm-serve`](skills/jetson-llm-serve), [`jetson-memory-audit`](skills/jetson-memory-audit), [`jetson-package`](skills/jetson-package), [`jetson-print-device-info`](skills/jetson-print-device-info), [`jetson-speculative-decoding`](skills/jetson-speculative-decoding) |
+| **Jetson Device** | Device-side agent skills for working with a live NVIDIA Jetson after boot — diagnostics, memory auditing, headless setup, inference memory tuning, LLM serving and benchmarking, packaging guidance, and speculative decoding. | [`jetson-diagnostic`](skills/jetson-diagnostic), [`jetson-headless-mode`](skills/jetson-headless-mode), [`jetson-inference-mem-tune`](skills/jetson-inference-mem-tune), [`jetson-llm-benchmark`](skills/jetson-llm-benchmark), [`jetson-llm-serve`](skills/jetson-llm-serve), [`jetson-memory-audit`](skills/jetson-memory-audit), [`jetson-package`](skills/jetson-package), [`jetson-print-device-info`](skills/jetson-print-device-info), [`jetson-speculative-decoding`](skills/jetson-speculative-decoding), [`jetson-video-benchmark`](skills/jetson-video-benchmark), [`jetson-video-capability`](skills/jetson-video-capability), [`jetson-video-pipeline`](skills/jetson-video-pipeline), [`jetson-video-recipe`](skills/jetson-video-recipe), [`jetson-video-setup`](skills/jetson-video-setup) |
 | **Medical AI Skills** | Agent-ready medical AI skills built on MONAI for DICOM handling, NVIDIA-hosted medical imaging model workflows, segmentation, synthesis, and evidence-oriented evaluation. | [`dicom-metadata-extract`](skills/dicom-metadata-extract), [`dicom-series-preflight`](skills/dicom-series-preflight), [`dicom-series-to-volume`](skills/dicom-series-to-volume), [`nv-generate-ct-rflow`](skills/nv-generate-ct-rflow), [`nv-generate-mr`](skills/nv-generate-mr), [`nv-generate-mr-brain`](skills/nv-generate-mr-brain), [`nv-generate-mr-brain-finetune`](skills/nv-generate-mr-brain-finetune), [`nv-generate-vae-finetune`](skills/nv-generate-vae-finetune), [`nv-reason-cxr`](skills/nv-reason-cxr), [`nv-segment-ct`](skills/nv-segment-ct), [`nv-segment-ct-finetune`](skills/nv-segment-ct-finetune), [`nv-segment-ctmr`](skills/nv-segment-ctmr) |
 | **Megatron-Core** | Large-scale distributed training — model parallelism, pipeline parallelism, and mixed precision. | [`mcore-create-issue`](skills/mcore-create-issue), [`mcore-linting-and-formatting`](skills/mcore-linting-and-formatting), [`mcore-run-on-slurm`](skills/mcore-run-on-slurm), [`mcore-split-pr`](skills/mcore-split-pr), [`mcore-testing`](skills/mcore-testing) |
 | **NeMo AutoModel** | NeMo AutoModel - PyTorch-native distributed training for LLMs/VLMs with Hugging Face support, recipes, launchers, and validation workflows. | [`nemo-automodel-distributed-training`](skills/nemo-automodel-distributed-training), [`nemo-automodel-launcher-config`](skills/nemo-automodel-launcher-config), [`nemo-automodel-model-onboarding`](skills/nemo-automodel-model-onboarding), [`nemo-automodel-recipe-development`](skills/nemo-automodel-recipe-development) |

@@ -0,0 +1,98 @@
+# Skill Benchmark: jetson-video-benchmark
+
+> ✅ **Overall verdict: PASS — Recommended for publication**
+
+## Publication Recommendation
+
+Recommended for publication based on the completed evaluation evidence in this report.
+
+## Evaluation Metadata
+
+- Skill: `jetson-video-benchmark`
+- Evaluation date: 2026-08-10
+- Evaluator version: `1.1.2`
+- Agents: Claude Code (`aws/anthropic/bedrock-claude-opus-4-8`), Codex (`openai/openai/gpt-5.5`)
+- Tasks: 7 evaluation tasks (7 positive)
+- Dataset digest: `sha256:e2d80ff3bff2819a83593a6167c1fd9119caa0d8febe727d4a947e1b84baec88` (skill-evaluator-dataset-snapshot/1)
+- Attempts per task: 1
+- Environment: `k8s-sandbox`
+- Tier 3 evidence: required for publication
+
+Each task attempt ran in its own isolated sandbox pod.
+
+## What This Report Answers
+
+The three-tier evaluation checks whether the skill:
+
+- is safe to use;
+- produces correct answers;
+- is discovered and activated when needed;
+- helps the agent complete the user's goal and expected workflow; and
+- avoids wasted skill and tool usage.
+
+## Results at a Glance
+
+| Measure | Claude Code (Baseline → Skill Uplift) | Codex (Baseline → Skill Uplift) |
+|---|---:|---:|
+| Overall | 47% → 97% (+50 points) | 49% → 92% (+43 points) |
+| Security | 100% → 100% (±0 points) | 86% → 100% (+14 points) |
+| Correctness | 51% → 100% (+49 points) | 69% → 83% (+14 points) |
+| Discoverability | 28% → 100% (+72 points) | 48% → 94% (+46 points) |
+| Effectiveness | 36% → 84% (+48 points) | 32% → 90% (+58 points) |
+| Efficiency | 19% → 100% (+81 points) | 11% → 92% (+82 points) |
+
+**How to read this table:** baseline is the same task attempted without the target skill. Uplift is `skill score - baseline score`, shown in percentage points.
+
+Example: `47% → 92% (+45 points)` means the skill-assisted run scored 92%, 45 percentage points above its 47% no-skill baseline.
+
+## Tier Status
+
+| Tier | Purpose | Status | Evidence |
+|---|---|---|---|
+| Tier 1 | Static validation | **PASSED WITH OBSERVATIONS** | 1 validator(s); 1 finding(s) |
+| Tier 2 | Semantic deduplication | **NOT RUN** | No result was recorded |
+| Tier 3 | Live agent evaluation | **PASS** | 2 agent(s); 7 task(s) |
+
+## Findings and Observations
+
+<details>
+<summary>Show detailed findings and successful checks</summary>
+
+- **MEDIUM** SCHEMA/body_recommended_section: Missing recommended section: '## Examples' (`skills/jetson-video-benchmark/SKILL.md`)
+
+</details>
+
+## Scoring Methodology
+
+<details>
+<summary>Show dimension definitions, source signals, and thresholds</summary>
+
+| Dimension | Question | Scored signals |
+|---|---|---|
+| Security | Is it safe to use? | `security` (100%) |
+| Correctness | Is the answer correct? | `accuracy` (100%) |
+| Discoverability | Was the right skill loaded when needed? | `skill_execution` (100%) |
+| Effectiveness | Did the skill help complete the task? | `goal_accuracy` (50%) + `behavior_check` (50%) |
+| Efficiency | Did it avoid wasted tool or skill usage? | `skill_efficiency` (100%) |
+
+- Dimension bands: PASS at 50% or above; NEUTRAL from 40% to below 50%; FAIL below 40%.
+- Overall Tier 3 lift: PASS at +5 points or more; FAIL at -10 points or less; values between those bands are NEUTRAL.
+- Overall verdict: PASS only when every configured dimension passes for at least one supported agent. Lift is reported as diagnostic evidence and does not override this gate.
+- The 50% attempt pass threshold is a separate per-task gate; it is not the dimension pass threshold.
+- Effectiveness is the equal-weight mean of goal completion (`goal_accuracy`) and expected workflow adherence (`behavior_check`).
+- Token efficiency is a separate report-only signal. It does not change a dimension score or the overall verdict.
+
+Signals present in this run:
+
+- `security` (Security): unsafe operations, secret leakage, and unauthorized access.
+- `skill_execution` (Skill Execution): whether the expected skill was found and executed.
+- `skill_efficiency` (Efficiency): routing quality, workspace-aware skill reads, and productive tool use.
+- `accuracy` (Accuracy): final-answer correctness against the reference answer.
+- `goal_accuracy` (Goal Accuracy): whether the user's goal was achieved.
+- `behavior_check` (Behavior Check): whether the expected workflow behavior was followed.
+
+</details>
+
+## Freshness
+
+Regenerate this benchmark when the skill, evaluation dataset, target agent/model, evaluator version, environment, or scoring policy changes.