NVIDIA AI Blueprint for video search and summarization (VSS) is a GPU-accelerated reference architecture for building video analytics agents with real-time verified alerts, visual Q&A, and automated reporting. The VSS Blueprint uses vision language models (VLMs) such as NVIDIA Cosmos, LLMs such as NVIDIA Nemotron, RAG, and NVIDIA NIMs.
NVIDIA AI Blueprint for Video Search and Summarization (VSS) is a reference architecture for applications that use AI to search, analyze, question, and summarize live or recorded video. It is intended for building video analytics agents with vision-language models, language models, retrieval-augmented generation, and NVIDIA NIM microservices. The catalogue entries provide coding-agent rules, skills, commands, agents, instructions, and settings associated with the project.
This repository also configures its own agents. See what video-search-and-summarization tells them →
NVIDIA-AI-Blueprints/video-search-and-summarization
Command Claude Code
Write BDD tests (feature file + pytest-bdd step implementations) for a new VIOS feature.
NVIDIA-AI-Blueprints/video-search-and-summarization
Command Claude Code
Get the exact build.sh command for your VIOS build scenario.
NVIDIA-AI-Blueprints/video-search-and-summarization
Command Claude Code
Scaffold a new VIOS VMS adaptor following the project's adaptor pattern.
NVIDIA-AI-Blueprints/video-search-and-summarization
Command Claude Code
Data-preserving git rebase and conflict resolution onto v2.1 (or any target branch).
NVIDIA-AI-Blueprints/video-search-and-summarization
Command Claude Code
Fetch all review comments on a GitLab MR, verify each one against the codebase, apply fixes for valid issues, and post a reply to every comment.
NVIDIA-AI-Blueprints/video-search-and-summarization
Command Claude Code
Review staged or recently changed VIOS code for Checkmarx and nspect security issues before committing.
NVIDIA-AI-Blueprints/video-search-and-summarization
Command Claude Code
Install MaaS MCP servers into Claude Code settings for GitLab, Outlook, Jira, Confluence, and other NVIDIA internal services.
NVIDIA-AI-Blueprints/video-search-and-summarization
Command Claude Code
Run VIOS UI tests using the playwright plugin.
NVIDIA-AI-Blueprints/video-search-and-summarization
Command Claude Code
Explain the VIOS stream-processor and sensor microservice deployment architecture: containers, ingress, RTSP, file-upload, WebRTC, Envoy, SDR, routing, and event flow.
NVIDIA-AI-Blueprints/video-search-and-summarization
Command Claude Code
VIOS deployment agent — build, deploy, stop, and manage the VIOS Docker Compose stack.
NVIDIA-AI-Blueprints/video-search-and-summarization
Command Claude Code
Execute VIOS/VST sanity test cases from the viosdevsanity.xlsx spreadsheet against a running VIOS web UI instance using Playwright. Tests are loaded dynamically from the xlsx — no hardcoded test matrix.
NVIDIA-AI-Blueprints/video-search-and-summarization
Command Claude Code
VIOS git conventions: branch names, commit messages, and merge-request descriptions. Use before creating any branch, commit, or MR.
NVIDIA-AI-Blueprints/video-search-and-summarization
Command Claude Code
Generate a complete VIOS multi-arch release note for a given build drop version. Fetches container images from GitLab compose.env, GitLab commit titles since the previous release tag, Jira VST tickets, NVBugs entries, Slack channel evidence, and Outlook email evidence, then renders and saves a formatted release note.
NVIDIA-AI-Blueprints/video-search-and-summarization
Command Claude Code
VIOS SQA agent — build containers, deploy, run BDD tests, and validate web UI.
NVIDIA-AI-Blueprints/video-search-and-summarization
Command Claude Code
Configure backend IP in config.tsx, install deps, and start the UI dev server.
NVIDIA-AI-Blueprints/video-search-and-summarization
Command Claude Code
Build the VST UI and deploy the static files into the vios tree (services/vios of the video-search-and-summarization repo), both ingress/vst-ui and webroot, then commit.
NVIDIA-AI-Blueprints/video-search-and-summarization
Command Claude Code
This skill should be used when the user asks to "add a feature", "fix a bug", "create a component", "implement UI", "update the VST UI", "change the dashboard", "modify the video player", or any other development task on the VST web client (vst-ui-ts TypeScript/React codebase). Covers the full development loop: plan →…
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: