What does this skill do, and when should you use it?
This skill guides standalone deployment of the vss-video-analytics-api REST service rather than the complete VSS warehouse blueprint. It covers configuration-source selection, data-log binding, required Elasticsearch connectivity, and optional Kafka integration. It also routes users through container startup, health verification, REST endpoint references, and troubleshooting. Deployment requires a VSS repository checkout, NGC credentials, Docker, and reachable Elasticsearch.
Guides installation through the NVIDIA skills CLI; checks the VSS repository path, NGC_CLI_API_KEY, Docker versions, and Elasticsearch reachability; prepares the data-log directory and runs Docker Compose; verifies the service with /livez, /docs, or /health; explains optional Kafka configuration and the Kafka-dependent dynamic configuration, dynamic calibration, and RTLS/AMR features; and diagnoses connection refusal, NGC authorization failures, and container OOM or model-loading errors.
- VSS operators who need only the video analytics REST API instead of the full warehouse stack.
- Deployment engineers who need a custom video-analytics-api configuration or alternate Elasticsearch or Kafka endpoints.
- Users who want a broker-less API deployment while accepting that Kafka-dependent features will not work.
- VSS operators who need to manage configuration or calibration through REST, or expose RTLS/AMR data through the API.
How do you install this skill?
- Confirm Docker Compose actions, host-network exposure, Elasticsearch/Kafka connectivity, and data-directory permissions before execution; do not use a long-lived NGC key without review.
- The service-shipped configuration enables Kafka by default; broker-less deployments should explicitly use an empty brokers configuration or Kafka-dependent features may fail.
- This assessment executed nothing and did not verify NGC image access, Docker versions, external documentation, or mainland-China network reachability.
- Shell / CLI
- Network access
- Local filesystem
Docker Engine 28.3.3Docker Compose plugin v2.39.1+ElasticsearchNGC_CLI_API_KEY
Install through the NVIDIA skills CLI: npx skills add nvidia/skills --skill vss-setup-video-analytics-api --yes. Add --agent codex or --agent claude-code to target a client. The README does not specify a manual destination; the CLI prompts for the installation target.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/NVIDIA/skills.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/skills/vss-setup-video-analytics-api" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
Once installed, send your agent any of these to trigger it:
- deploy video analytics api
- run video-analytics-api standalone
Prompt a compatible agent with “deploy video analytics api” or “run video-analytics-api standalone”. The documented minimal flow is to point VSS_APPS_DIR to the repository’s deploy/docker directory, load NGC_CLI_API_KEY securely, prepare a writable VSS_DATA_DIR, confirm Docker Engine 28.3.3, Compose 2.39.1+, and Elasticsearch, then run docker compose -f services/analytics/video-analytics-api/compose.yml up -d vss-video-analytics-api and verify with curl -sf http://localhost:8081/livez. The full workflow, configuration fields, endpoint table, and dependency notes are in references/.
What are this skill's strengths and limitations?
- Clearly scoped to standalone video-analytics-api deployment rather than the full warehouse stack.
- Covers configuration sources, data-log binding, Elasticsearch, optional Kafka, verification, and troubleshooting.
- Includes copyable Docker Compose and health-check commands.
- Documents Kafka runtime dependencies for dynamic configuration, dynamic calibration, and RTLS/AMR.
- Requires the stated Docker Engine and Compose versions and valid NGC credentials.
- Elasticsearch is mandatory; Kafka is optional but required for several features.
- A matching VSS profile or microservice must be deployed and reachable by the caller.
- The supplied material provides no platform matrix, automated test results, or performance benchmarks.
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| VSS Standalone Video Analytics API this page ✓ NVIDIA · Official | 54 · Use with care | ★ 3.6k | 13d ago | Apache-2.0 |
| RT-VLM Dense Captioning Deployment ✓ NVIDIA · Official | 45 · Not recommended | ★ 3.6k | 13d ago | Apache-2.0 |
| VSS Video Embedding Deployment ✓ NVIDIA · Official | 57 · Use with care | ★ 3.6k | 13d ago | Apache-2.0 |
| VSS Standalone Behavior Analytics ✓ NVIDIA · Official | 51 · Use with care | ★ 3.6k | 13d ago | Apache-2.0 |
| RTVI-CV 2D Detection & Tracking Operations ✓ NVIDIA · Official | 54 · Use with care | ★ 3.6k | 13d ago | Apache-2.0 |
Unlike vss-deploy-profile for the warehouse or alerts profiles, this skill deploys only the vss-video-analytics-api container and excludes the UI, agent, and perception stack. Behavior creation and incident detection are routed to vss-setup-behavior-analytics.
How did FollowSkills review this skill?
The documentation explicitly requires interactive handling of NGC_CLI_API_KEY, forbids committing or logging it, and discloses Elasticsearch, Kafka, NGC, and Docker data flows. However, host networking, a long-lived credential with broad NGC access, and the lack of explicit pre-execution confirmation, least-privilege guidance, and complete rollback procedures warrant deductions.
Routing, prerequisites, configuration choices, health checks, and troubleshooting are generally consistent. Static review did not execute the key path, test coverage is limited, and Kafka configuration can still produce runtime failures in broker-less environments, so the score remains below the static ceiling.
Trigger phrases, target scenarios, and the boundary against full warehouse deployment are clear; Elasticsearch is identified as required and Kafka as optional. Chinese interaction and mainland-China reachability are not addressed, and NGC and related documentation availability may vary by environment, so points are deducted.
The material is reasonably layered and includes configuration, deployment, troubleshooting, limitations, version 3.2.0, Apache-2.0 licensing, and repository provenance. However, the benchmark reports deeply nested references and an invalid author format, while changelog coverage, named maintenance responsibility, and an explicit update path are missing; the supplied repository license metadata is also NOASSERTION.
The skill supplies directly usable Docker Compose commands, configuration paths, dependency startup guidance, and livez verification, supported by one positive evaluation report. That report covers only one task and was not reproduced in this review; results still depend on user credentials, host state, and external services, limiting confidence in completeness.
Committed evaluation manifests, expected behaviors, benchmark results, and static-validation findings provide some auditability. However, coverage is one positive task with no negative-trigger set, and no CI or comprehensive test suite can be confirmed from the supplied material, so evidence remains thin.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →