Generated banner for Rtvi Cv Customize Model
Generated logo for Rtvi Cv Customize Model

Rtvi Cv Customize Model

NVIDIAOfficialCreated by NVIDIA CORPORATION

  • Skill
  • Workflow
  • Software development
  • Apache-2.0
  • Medium risk

From the author

How to swap the DeepStream CV detection model in the VSS Alerts Blueprint verification (2d_cv) mode - covers ONNX export, custom bbox parsers, compose mount gotchas, nvinfer config, runtime TRT engine build, deployment, and a segmentation-capable model addendum handoff.
— the description by NVIDIA CORPORATION

AutoGineer summary

Written by AutoGineer from the skill’s own files, not by the author

This skill provides a comprehensive procedure for swapping the primary GPU inference engine (GIE) in the VSS Alerts Blueprint verification (2d_cv) mode. It covers the end-to-end process of preparing a custom ONNX model, configuring the DeepStream nvinfer engine, building a custom bounding box parser, and updating the runtime environment to ensure the new model is correctly deployed and executed within the VSS perception container.

What it does

  • Guides the export and staging of custom ONNX detection models.
  • Details the configuration of DeepStream nvinfer properties for custom models.
  • Explains the creation and integration of custom C++ bounding box parsers.
  • Provides instructions for runtime TensorRT engine generation and caching.
  • Outlines necessary modifications to Docker compose and startup scripts for deployment.
  • Offers troubleshooting steps for common parser and model loading failures.

When to use it

  • When you need to replace the default detector in the VSS Alerts Blueprint verification mode.
  • When debugging issues with ONNX model staging or missing TensorRT engine builds.
  • When resolving DeepStream parser load failures or symbol mismatches.
  • When you need to integrate a custom detection model into an existing VSS v3.2.1 deployment.

Advantages

  • Enables model flexibility within the VSS perception pipeline.
  • Provides a structured approach to handling custom inference configurations.
  • Includes specific guidance on avoiding common pitfalls like ghost directory mounts.
  • Ensures compatibility with the VSS Alerts Blueprint architecture.

Usage examples

“Replace the VSS Alerts Blueprint verification detector with a YOLO11 ONNX model and redeploy perception-alerts.”
The user receives a step-by-step guide to staging the ONNX file, updating the nvinfer config, and modifying the startup script to use the new model.
“DeepStream says dlsym failed for the bbox parser after loading the custom parser library. What should I check?”
The agent identifies the need to verify that the parser symbol name matches exactly between the C++ code, the nvinfer config, and the macro definition.
“I mounted the ONNX file path directly in compose before the host file existed. Is that okay?”
The agent explains that this creates a ghost directory and advises mounting the parent directory instead.

Install

  1. Download the skill as a ZIP:

    https://autogineer.com/agent-skills/nvidia/rtvi-cv-customize-model/download
  2. Unzip it into the cross-client skills folder of your project:

    mkdir -p .agents/skills && unzip rtvi-cv-customize-model.zip -d .agents/skills

Documentation by NVIDIA CORPORATION

CV Detection Model Customization — VSS Alerts Blueprint (2d_cv mode only)

The RT-CV perception container (vss-rt-cv) runs a DeepStream pipeline with a configurable primary GIE (GPU Inference Engine). By default it uses GDINO or RTDETR. This guide covers replacing it with any ONNX-format model, using YOLOv11 COCO 80 as the worked example.

This only applies to --mode verification (2d_cv). Real-time alerts mode (2d_vlm) has no CV detector.

For instance-segmentation or detection-plus-mask models, complete references/segmentation-model-contract.md before writing parser or handoff code.


When to use

Use this skill when the user wants to:

  • swap the stock vss-rt-cv detector for another ONNX model in verification mode,

  • debug a broken ONNX staging path, ghost-directory bind mount, or missing runtime TRT engine build,

  • fix a DeepStream parser load failure such as dlsym failed on the bbox parser symbol.

Do not use this skill for:

  • 2d_vlm real-time alerts mode,

  • scaffolding a brand-new standalone RTVI CV microservice (use rtvi-cv-scaffold-vss-service).

Instructions

  • Keep the answer scoped to VSS Alerts Blueprint verification mode (2d_cv) unless the user explicitly asks to compare modes.

  • Paths beginning with deploy/docker/ are relative to the VSS Blueprint repository, not the DeepStream repository. Clone or reuse a VSS checkout that is v3.2.1 or compatible, then run these commands from that repository root (see VSS Quickstart).

  • If the user asks about ONNX staging or a compose mount, explicitly say that mounting a missing file path is wrong: use the stock parent-directory mount, keep the ONNX under ${VSS_DATA_DIR}/models/yolo, and re-stage the ONNX file after any dev-profile.sh up that recreates the models directory.

  • If the user asks about parser load failures, line up parse-bbox-func-name, the extern "C" function symbol, and CHECK_CUSTOM_PARSE_FUNC_PROTOTYPE(...), then rebuild the .so with both DeepStream and CUDA include paths present.

  • If the model is detection-plus-mask or instance segmentation, route the mask-contract decision first through references/segmentation-model-contract.md before editing parser or handoff logic.

Examples

  • "Replace the VSS Alerts Blueprint verification detector with a YOLO11 ONNX model and redeploy perception-alerts."

  • "DeepStream says dlsym failed for the bbox parser after loading the custom parser library. What should I check?"

  • "I mounted the ONNX file path directly in compose before the host file existed. Is that okay?"


Architecture

NVStreamer (RTSP) → SDR (port 9010) → DeepStream (perception-alerts)
                                            └─ primary-gie (nvinfer)
                                                  ├─ ONNX → TRT engine (built once, then cached)
                                                  ├─ custom bbox parser (.so)
                                                  └─ label file (.txt)
                                            └─ Kafka → mdx-raw
                                                  └─ vss-behavior-analytics
                                                        └─ mdx-incidents
                                                              └─ vlm-as-verifier

VSS source location

This skill is documentation-only and does not ship the VSS deployment sources. Use a VSS v3.2.1 or compatible checkout; run all paths and commands from that repository root. Clone/LFS setup, stock Alerts profile files, and the YOLOv11 customization file tree are in references/vss-source-layout.md.


Adapting to a Different Model

YOLOv11 COCO 80 is the worked example throughout all steps below. The same pattern applies to any ONNX-format detector — substitute at these points:

StepWhat to change
Step 1Replace the export procedure with whatever your model's training library requires. Confirm the resulting .onnx exists on the host before continuing.
Steps 2–3Inspect your model's actual output tensor name, shape, layout, and whether NMS is applied in-graph. Do not assume it matches YOLOv11. The deepstream-dev skill (skills/deepstream-dev/) has a generation-by-generation YOLO output format table and references/nvinfer_config.md for full property reference.
Step 4Update output-blob-names to your tensor name, infer-dims to your input shape, and cluster-mode to match whether NMS is in-graph (4) or not (2).
Step 5Update the --onnx path, the input tensor name in --minShapes/--optShapes/--maxShapes (not images for non-YOLO models), and the sed commands to reference your nvinfer .txt file. Add a new if [[ $MODEL_NAME_2D == "YOURMODEL" ]] block rather than editing the YOLO block.

Step 1: Stage Your ONNX Model (Host, One-Time)

Obtain a TensorRT-compatible ONNX model and place it at ${VSS_DATA_DIR}/models/yolo/yolo11s.onnx before continuing.

See references/yolov11-onnx-export.md for the export settings and tensor layout used in this reference. Adapt the parser and nvinfer config in later steps to match your model's actual output.

After every full dev-profile.sh up, restore ownership of the recreated models/yolo directory and re-stage your ONNX file:

sudo mkdir -p "${VSS_DATA_DIR}/models/yolo"
sudo chown -R $(id -u):$(id -g) "${VSS_DATA_DIR}/models/yolo"
# Copy your ONNX model to ${VSS_DATA_DIR}/models/yolo/yolo11s.onnx

Steps 2–3: Inspect the Model Output and Build the Parser

Read and follow references/yolov11-parser.md before creating ${VSS_PROFILE_DIR}/deepstream/custom_parser/nvdsparseyolov11.cpp or editing the parser build in Dockerfiles/perception.Dockerfile.

Before writing any parser code, confirm these four things for your model:

  • Output tensor name and shape — drives output-blob-names and infer-dims in Step 4

  • Pre-NMS or post-NMS — drives cluster-mode in Step 4; see deepstream-dev Rule 13 for the generation-by-generation breakdown

  • Coordinate semantics — cx/cy/w/h center-format or x1/y1/x2/y2 corner-format; the YOLOv11 example uses center-format (see references/yolov11-parser.md)

  • Parser symbol name — the extern "C" function name must match parse-bbox-func-name in Step 4 and CHECK_CUSTOM_PARSE_FUNC_PROTOTYPE(...) exactly

Confirm the model output tensor layout instead of assuming it matches YOLOv11. Keep the parser symbol identical in the exported C++ function, CHECK_CUSTOM_PARSE_FUNC_PROTOTYPE(...), and the Step 4 parse-bbox-func-name setting.

For instance segmentation or detection-plus-mask models, complete the contract decisions table in references/segmentation-model-contract.md before writing parser or handoff code.


Step 4: nvinfer Config

${VSS_PROFILE_DIR}/deepstream/configs/yolov11.txt — the primary GIE sub-config referenced by the DeepStream run config.

model-engine-file and batch-size are patched at container startup by the ds-start.sh block in Step 5. Copy the values below as-is; Step 5 will overwrite both with the correct engine path and sensor count before DeepStream reads the file.

[property]
gpu-id=0
net-scale-factor=0.0039215697906911373    # 1/255 input normalization
model-engine-file=/opt/storage/yolo11s_fp16.engine
onnx-file=/opt/storage/yolo/yolo11s.onnx
batch-size=1                             # patched at startup from NUM_SENSORS
network-mode=2                           # 0=FP32, 1=INT8, 2=FP16
network-type=0                           # 0=Detector
num-detected-classes=80
interval=0
gie-unique-id=1
output-blob-names=output0               # must match your model's output tensor name
infer-dims=3;640;640                    # C;H;W
maintain-aspect-ratio=1
parse-bbox-func-name=NvDsInferParseCustomYoloE  # must match extern "C" function name in .so
custom-lib-path=/opt/deepstream-yolo/libnvdsparseyolov11.so
labelfile-path=/opt/nvidia/deepstream/deepstream/sources/apps/sample_apps/metropolis_perception_app/mounted-configs/yolo-coco-labels.txt
cluster-mode=2                   # 2=NMS — required: Ultralytics ONNX has no NMS in-graph

[class-attrs-all]
pre-cluster-threshold=0.25
topk=300
nms-iou-threshold=0.45

Step 5: Runtime TRT Engine Build (ds-start.sh)

TRT engines are GPU-architecture-specific. Build the engine on first startup when it is missing, then reuse it on subsequent restarts. Rebuild an existing engine only when FORCE_REBUILD=true. The stock Alerts compose file bind-mounts ${VSS_DATA_DIR}/models/ at /opt/storage/, so the generated engine persists on the host.

Which ds-start.sh? Stock Alerts runs services/rtvi/rtvi-cv/ds-start.sh (via extends), not ${VSS_PROFILE_DIR}/deepstream/init-scripts/ds-start.sh. This skill edits the profile copy, then remounts it in Step 6 so those edits actually run. Full map of the three paths, remount trade-offs, and DS_CONFIG_FILE pairing: references/ds-start-entrypoint.md.

Edit ${VSS_PROFILE_DIR}/deepstream/init-scripts/ds-start.sh, then override the stock rtvi-cv ds-start.sh bind mount in the Alerts profile compose file as shown in Step 6. Without that override, profile-script edits never execute.

Critical — honor DS_CONFIG_FILE. The profile script sets CONFIG_FILE=${1:-...} and never reads DS_CONFIG_FILE. The Alerts compose command invokes ds-start.sh with no positional args, while Step 6 sets DS_CONFIG_FILE to the absolute mounted-configs run config (not the stock .../configs/... path). Replace the script's existing CONFIG_FILE=... assignment with:

# Prefer $1 when provided; otherwise use DS_CONFIG_FILE from compose (absolute path).
# Without this, YOLO_CONFIG resolves to ./yolov11.txt and the stock detector keeps running.
CONFIG_FILE="${1:-${DS_CONFIG_FILE:-/opt/nvidia/deepstream/deepstream/sources/apps/sample_apps/metropolis_perception_app/mounted-configs/run_config-api-rtdetr-protobuf.txt}}"

Then add the YOLO block (before the GDINO/RT-DETR model branches is fine):

if [[ $MODEL_NAME_2D == "YOLO" ]]; then
    YOLO_ONNX=/opt/storage/yolo/yolo11s.onnx
    YOLO_ENGINE=/opt/storage/yolo11s_fp16.engine
    YOLO_ENGINE_TMP="${YOLO_ENGINE}.building"
    FORCE_REBUILD=${FORCE_REBUILD:-false}

    if [[ ! -f "$YOLO_ONNX" ]]; then
        echo "ERROR: ONNX not found at ${YOLO_ONNX}. Stage your model there before starting."
        exit 1
    fi

    if [[ ! -s "$YOLO_ENGINE" || "${FORCE_REBUILD,,}" == "true" ]]; then
        echo "Building TensorRT engine: ${YOLO_ENGINE}"
        rm -f "$YOLO_ENGINE_TMP"
        if /usr/src/tensorrt/bin/trtexec \
            --onnx=${YOLO_ONNX} \
            --minShapes=images:1x3x640x640 \
            --optShapes=images:${NUM_SENSORS}x3x640x640 \
            --maxShapes=images:${NUM_SENSORS}x3x640x640 \
            --fp16 --saveEngine=${YOLO_ENGINE_TMP}; then
            mv "$YOLO_ENGINE_TMP" "$YOLO_ENGINE"
        else
            rm -f "$YOLO_ENGINE_TMP"
            echo "ERROR: TensorRT engine build failed; existing engine was preserved."
            exit 1
        fi
    else
        echo "Reusing cached TensorRT engine: ${YOLO_ENGINE}"
    fi

    # Patch the run config to use yolov11.txt for the primary GIE
    # (relative name is intentional: DeepStream resolves it against CONFIG_FILE's dir)
    sed -i '/^\[primary-gie\]/,/^\[/{s/config-file=.*/config-file=yolov11.txt/;}' "$CONFIG_FILE"
    # Patch engine path and batch size into yolov11.txt (same dir as CONFIG_FILE)
    YOLO_CONFIG="$(dirname "${CONFIG_FILE}")/yolov11.txt"
    sed -i "s|model-engine-file=.*|model-engine-file=${YOLO_ENGINE}|" "${YOLO_CONFIG}"
    sed -i "/^\[property\]/,/^\[/{s/^batch-size=.*/batch-size=${NUM_SENSORS}/;}" "${YOLO_CONFIG}"
fi

To add a different model, add a new if [[ $MODEL_NAME_2D == "YOURMODEL" ]] block. Update:

  • --onnx path and --saveEngine output path

  • --minShapes/--optShapes/--maxShapes input tensor name and dimensions for your model

  • The sed commands to point at your nvinfer .txt file


Step 6: Compose + Env

In the existing ${VSS_PROFILE_DIR}/compose.yml, update the perception-alerts service:

perception-alerts:
  build:
    context: $VSS_APPS_DIR/developer-profiles/dev-profile-alerts
    dockerfile: Dockerfiles/perception.Dockerfile
  volumes:
    # Keep the stock parent-directory mount. Do not replace it with a direct
    # mount of yolo11s.onnx: Docker creates a ghost directory if the host
    # file does not exist.
    - $VSS_DATA_DIR/models/:/opt/storage/
    - $VSS_APPS_DIR/developer-profiles/dev-profile-alerts/deepstream/configs/:/opt/nvidia/deepstream/deepstream/sources/apps/sample_apps/metropolis_perception_app/mounted-configs/
    # Override stock rtvi-cv ds-start.sh (extends bind). Without this,
    # edits under deepstream/init-scripts/ never run — see
    # references/ds-start-entrypoint.md.
    - $VSS_APPS_DIR/developer-profiles/dev-profile-alerts/deepstream/init-scripts/ds-start.sh:/opt/nvidia/deepstream/deepstream/sources/apps/sample_apps/metropolis_perception_app/ds-start.sh:ro
  environment:
    MODEL_NAME_2D: ${MODEL_NAME_2D}
    NUM_SENSORS: ${NUM_SENSORS}
    FORCE_REBUILD: ${FORCE_REBUILD:-false}
    # Use mounted-configs (where yolov11.txt lives), not stock .../configs/...
    DS_CONFIG_FILE: /opt/nvidia/deepstream/deepstream/sources/apps/sample_apps/metropolis_perception_app/mounted-configs/run_config-api-rtdetr-protobuf.txt

Set in ${VSS_PROFILE_DIR}/.env:

MODEL_NAME_2D="YOLO"
NUM_SENSORS=1
FORCE_REBUILD=false

Keep FORCE_REBUILD=false for normal starts. Compose's --force-recreate option recreates the container but does not rebuild the TensorRT engine unless FORCE_REBUILD=true.

Gotcha — generated.env wins at restart. Step 7's redeploy uses --env-file .../generated.env, which is produced by dev-profile.sh from .env. Editing .env alone does not change FORCE_REBUILD for that compose command. For a one-shot engine rebuild, pass the override on the shell (see Step 7). To persist the value for later dev-profile.sh runs, edit .env and re-run dev-profile.sh so it regenerates generated.env.


Step 7: Deploy

Run from the video-search-and-summarization repository root (VSS_ROOT / VSS_DEPLOY_DIR / VSS_PROFILE_DIR from VSS source location):

: "${VSS_ROOT:=$PWD}"
: "${VSS_DEPLOY_DIR:=${VSS_ROOT}/deploy/docker}"
: "${VSS_PROFILE_DIR:=${VSS_DEPLOY_DIR}/developer-profiles/dev-profile-alerts}"

1. Single-GPU host (only GPU 0): edit ${VSS_PROFILE_DIR}/.env so these keys match the required final values below. Edit the source profile env file, not generated.env (dev-profile.sh regenerates that). Stock Alerts reserves GPU 0 and puts RT-VLM / LLM / VLM on GPU 1; dev-profile.sh CLI flags cannot clear RESERVED_DEVICE_IDS or set FIXED_SHARED_DEVICE_IDS. Confirm device IDs exist with nvidia-smi --query-gpu=index --format=csv,noheader,nounits before deploying.

RESERVED_DEVICE_IDS=''
FIXED_SHARED_DEVICE_IDS='0'
RT_CV_DEVICE_ID='0'
RT_VLM_DEVICE_ID='0'
LLM_DEVICE_ID='0'
VLM_DEVICE_ID='0'

2. Deploy the official Alerts verification profile. Select the hardware profile supported by your host, as documented by the VSS Quickstart. Do not present the remaining commands as a copy-paste handoff unless execution is blocked.

"${VSS_DEPLOY_DIR}/scripts/dev-profile.sh" up \
  -p alerts \
  -m verification \
  -H <H100|L40S|RTXPRO4500BW|RTXPRO6000BW|DGX-SPARK|IGX-THOR|AGX-THOR|OTHER>

# 3. dev-profile.sh recreates the model directory. Restore ownership and
# re-stage your ONNX model.
export VSS_DATA_DIR="${VSS_DEPLOY_DIR}/data-dir"
sudo mkdir -p "${VSS_DATA_DIR}/models/yolo"
sudo chown -R $(id -u):$(id -g) "${VSS_DATA_DIR}/models/yolo"
# Copy your ONNX model to ${VSS_DATA_DIR}/models/yolo/yolo11s.onnx

# 4. Rebuild and recreate the customized perception service using the compose
# project generated by dev-profile.sh.
cd "${VSS_DEPLOY_DIR}"
docker compose \
  --env-file developer-profiles/dev-profile-alerts/generated.env \
  up -d --build --force-recreate perception-alerts

# One-shot TensorRT rebuild (after ONNX / shape / GPU / TRT changes).
# Inline FORCE_REBUILD overrides generated.env for this invocation only:
# FORCE_REBUILD=true docker compose \
#   --env-file developer-profiles/dev-profile-alerts/generated.env \
#   up -d --build --force-recreate perception-alerts

TRT engine build takes approximately 15–30 seconds on Blackwell hardware. Subsequent restarts reuse the engine persisted under ${VSS_DATA_DIR}/models/.


Common Gotchas

See references/common-gotchas.md for single-GPU reservation failures, wipe/restage after dev-profile.sh up, ghost file mounts, dlsym parser mismatches, and Redis stream contamination.

Files

LICENSE

Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.


                                 Apache License
                           Version 2.0, January 2004
                        http://www.apache.org/licenses/

   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION

   1. Definitions.

      "License" shall mean the terms and conditions for use, reproduction,
      and distribution as defined by Sections 1 through 9 of this document.

      "Licensor" shall mean the copyright owner or entity authorized by
      the copyright owner that is granting the License.

      "Legal Entity" shall mean the union of the acting entity and all
      other entities that control, are controlled by, or are under common
      control with that entity. For the purposes of this definition,
      "control" means (i) the power, direct or indirect, to cause the
      direction or management of such entity, whether by contract or
      otherwise, or (ii) ownership of fifty percent (50%) or more of the
      outstanding shares, or (iii) beneficial ownership of such entity.

      "You" (or "Your") shall mean an individual or Legal Entity
      exercising permissions granted by this License.

      "Source" form shall mean the preferred form for making modifications,
      including but not limited to software source code, documentation
      source, and configuration files.

      "Object" form shall mean any form resulting from mechanical
      transformation or translation of a Source form, including but
      not limited to compiled object code, generated documentation,
      and conversions to other media types.

      "Work" shall mean the work of authorship, whether in Source or
      Object form, made available under the License, as indicated by a
      copyright notice that is included in or attached to the work
      (an example is provided in the Appendix below).

      "Derivative Works" shall mean any work, whether in Source or Object
      form, that is based on (or derived from) the Work and for which the
      editorial revisions, annotations, elaborations, or other modifications
      represent, as a whole, an original work of authorship. For the purposes
      of this License, Derivative Works shall not include works that remain
      separable from, or merely link (or bind by name) to the interfaces of,
      the Work and Derivative Works thereof.

      "Contribution" shall mean any work of authorship, including
      the original version of the Work and any modifications or additions
      to that Work or Derivative Works thereof, that is intentionally
      submitted to Licensor for inclusion in the Work by the copyright owner
      or by an individual or Legal Entity authorized to submit on behalf of
      the copyright owner. For the purposes of this definition, "submitted"
      means any form of electronic, verbal, or written communication sent
      to the Licensor or its representatives, including but not limited to
      communication on electronic mailing lists, source code control systems,
      and issue tracking systems that are managed by, or on behalf of, the
      Licensor for the purpose of discussing and improving the Work, but
      excluding communication that is conspicuously marked or otherwise
      designated in writing by the copyright owner as "Not a Contribution."

      "Contributor" shall mean Licensor and any individual or Legal Entity
      on behalf of whom a Contribution has been received by Licensor and
      subsequently incorporated within the Work.

   2. Grant of Copyright License. Subject to the terms and conditions of
      this License, each Contributor hereby grants to You a perpetual,
      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
      copyright license to reproduce, prepare Derivative Works of,
      publicly display, publicly perform, sublicense, and distribute the
      Work and such Derivative Works in Source or Object form.

   3. Grant of Patent License. Subject to the terms and conditions of
      this License, each Contributor hereby grants to You a perpetual,
      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
      (except as stated in this section) patent license to make, have made,
      use, offer to sell, sell, import, and otherwise transfer the Work,
      where such license applies only to those patent claims licensable
      by such Contributor that are necessarily infringed by their
      Contribution(s) alone or by combination of their Contribution(s)
      with the Work to which such Contribution(s) was submitted. If You
      institute patent litigation against any entity (including a
      cross-claim or counterclaim in a lawsuit) alleging that the Work
      or a Contribution incorporated within the Work constitutes direct
      or contributory patent infringement, then any patent licenses
      granted to You under this License for that Work shall terminate
      as of the date such litigation is filed.

   4. Redistribution. You may reproduce and distribute copies of the
      Work or Derivative Works thereof in any medium, with or without
      modifications, and in Source or Object form, provided that You
      meet the following conditions:

      (a) You must give any other recipients of the Work or
          Derivative Works a copy of this License; and

      (b) You must cause any modified files to carry prominent notices
          stating that You changed the files; and

      (c) You must retain, in the Source form of any Derivative Works
          that You distribute, all copyright, patent, trademark, and
          attribution notices from the Source form of the Work,
          excluding those notices that do not pertain to any part of
          the Derivative Works; and

      (d) If the Work includes a "NOTICE" text file as part of its
          distribution, then any Derivative Works that You distribute must
          include a readable copy of the attribution notices contained
          within such NOTICE file, excluding those notices that do not
          pertain to any part of the Derivative Works, in at least one
          of the following places: within a NOTICE text file distributed
          as part of the Derivative Works; within the Source form or
          documentation, if provided along with the Derivative Works; or,
          within a display generated by the Derivative Works, if and
          wherever such third-party notices normally appear. The contents
          of the NOTICE file are for informational purposes only and
          do not modify the License. You may add Your own attribution
          notices within Derivative Works that You distribute, alongside
          or as an addendum to the NOTICE text from the Work, provided
          that such additional attribution notices cannot be construed
          as modifying the License.

      You may add Your own copyright statement to Your modifications and
      may provide additional or different license terms and conditions
      for use, reproduction, or distribution of Your modifications, or
      for any such Derivative Works as a whole, provided Your use,
      reproduction, and distribution of the Work otherwise complies with
      the conditions stated in this License.

   5. Submission of Contributions. Unless You explicitly state otherwise,
      any Contribution intentionally submitted for inclusion in the Work
      by You to the Licensor shall be under the terms and conditions of
      this License, without any additional terms or conditions.
      Notwithstanding the above, nothing herein shall supersede or modify
      the terms of any separate license agreement you may have executed
      with Licensor regarding such Contributions.

   6. Trademarks. This License does not grant permission to use the trade
      names, trademarks, service marks, or product names of the Licensor,
      except as required for reasonable and customary use in describing the
      origin of the Work and reproducing the content of the NOTICE file.

   7. Disclaimer of Warranty. Unless required by applicable law or
      agreed to in writing, Licensor provides the Work (and each
      Contributor provides its Contributions) on an "AS IS" BASIS,
      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
      implied, including, without limitation, any warranties or conditions
      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
      PARTICULAR PURPOSE. You are solely responsible for determining the
      appropriateness of using or redistributing the Work and assume any
      risks associated with Your exercise of permissions under this License.

   8. Limitation of Liability. In no event and under no legal theory,
      whether in tort (including negligence), contract, or otherwise,
      unless required by applicable law (such as deliberate and grossly
      negligent acts) or agreed to in writing, shall any Contributor be
      liable to You for damages, including any direct, indirect, special,
      incidental, or consequential damages of any character arising as a
      result of this License or out of the use or inability to use the
      Work (including but not limited to damages for loss of goodwill,
      work stoppage, computer failure or malfunction, or any and all
      other commercial damages or losses), even if such Contributor
      has been advised of the possibility of such damages.

   9. Accepting Warranty or Additional Liability. While redistributing
      the Work or Derivative Works thereof, You may choose to offer,
      and charge a fee for, acceptance of support, warranty, indemnity,
      or other liability obligations and/or rights consistent with this
      License. However, in accepting such obligations, You may act only
      on Your own behalf and on Your sole responsibility, not on behalf
      of any other Contributor, and only if You agree to indemnify,
      defend, and hold each Contributor harmless for any liability
      incurred by, or claims asserted against, such Contributor by reason
      of your accepting any such warranty or additional liability.

   END OF TERMS AND CONDITIONS

   APPENDIX: How to apply the Apache License to your work.

      To apply the Apache License to your work, attach the following
      boilerplate notice, with the fields enclosed by brackets "[]"
      replaced with your own identifying information. (Don't include
      the brackets!)  The text should be enclosed in the appropriate
      comment syntax for the file format. We also recommend that a
      file or class name and description of purpose be included on the
      same "printed page" as the copyright notice for easier
      identification within third-party archives.

   Copyright [yyyy] [name of copyright owner]

   Licensed under the Apache License, Version 2.0 (the "License");
   you may not use this file except in compliance with the License.
   You may obtain a copy of the License at

       http://www.apache.org/licenses/LICENSE-2.0

   Unless required by applicable law or agreed to in writing, software
   distributed under the License is distributed on an "AS IS" BASIS,
   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
   See the License for the specific language governing permissions and
   limitations under the License.

Requirements and compatibility

Works with: Claude Code, Claude apps, OpenAI Codex, GitHub Copilot, Cursor, Gemini CLI, Kiro, OpenCode, Goose, Amp, Roo Code, JetBrains Junie, Factory Droid, Devin (ex-Windsurf), Mistral Vibe, OpenHands, Letta

“Works with” means the tool reads the open Agent Skills format; a skill written for one tool may use features others lack.

Needs: Needs a command-line tool

Security

What our scanner noticed· Medium risk

  • Large base64-encoded blobs
  • Instructions that use the network

A static scan is not a guarantee. Skills are third-party content: review a skill before you let an agent use it.

Attribution and licence

Created by
NVIDIA CORPORATION
Publisher
NVIDIA (Official)
Source
NVIDIA/skills · skills/rtvi-cv-customize-model · commit fce3cca
Licence
Apache License 2.0 (redistribution allowed)
  • LICENSE (added by AutoGineer from LICENSE-APACHE)
  • ATTRIBUTION.md (written by AutoGineer)
Synced

AutoGineer is not affiliated with or endorsed by NVIDIA CORPORATION or NVIDIA.

Download ZIP

The ZIP holds the skill’s files, its licence files and an ATTRIBUTION.md. Download rtvi-cv-customize-model.zip