NVIDIA

Validate coding-sequence CSVs, extract public CodonFM/Encodon embeddings, and choose checkpoints for downstream property modeling.

Fine-tune public CodonFM Encodon checkpoints on labeled coding-sequence or coding-variant data using LoRA, head-only, or full fine-tuning. Use when a user explicitly asks to fine-tune CodonFM or Encodon for regression or classification. Support generic…

Validate, prepare, or run public CodonFM Encodon masked-codon variant scoring and review compatibility of its scoring workflows. Use only when the user explicitly requests CodonFM or Encodon, or that context is already established in the conversation. Do not…

Set up the public CodonFM v1 repository and download public Encodon checkpoints. Use for requests to build or launch the CodonFM development container, configure local data/checkpoint mounts, verify GPU access, or download public Encodon 80M, 600M, 1B, or…

Build, migrate, review, and maintain third-party NVIDIA NeMo Fabric adapters against the public adapter contract. Use when creating adapter or target descriptors, mapping AgentConfig into an agent harness or custom-agent runtime, implementing…

Use this skill when integrating NVIDIA NeMo Fabric into a consumer application, service, evaluation harness, or platform through the typed Python SDK — translating the consumer's own application, job, or deployment config into an in-memory FabricConfig,…

Model and publish semantic definitions in Auto Ontology. Use for terms, relationships, measures, imports, and governed results—not deployment or querying.

Query Auto Ontology and validate generated SQL, rows, and answers. Use for MCP or REST access, readiness, authentication, conversations, and grounded questions.

Add SimReady capability docs, requirement indexes, validation stubs, and registrations for new requirement families.

Use for adding SimReady feature adapters that mutate USD assets between exact feature or profile versions.

Use for adding new SimReady feature markdown pages and matching JSON manifests from nv_core/sr_specs/docs/guides/features/feature-template.md, including FET runtime naming, integer feature versions, requirement mapping, validation strategy, index entries, and…

Use for adding SimReady profile versions with feature bundles, docs, indexes, and validation notes.

Use for adding atomic SimReady requirements with stable IDs, docs, examples, indexes, and validator follow-up.

Use for adding or updating SimReady Foundation runtime tests in a tier-owned Benchmark test package, wiring the package through the tier descriptor and wheel, and keeping feature/profile documentation and test registration synchronized. Delegates generic…

Use for adding executable SimReady validators that report requirement IDs with focused pass/fail coverage.

Use for repairing exact FET_000_ISAAC SimReady conformance: Isaac packaging Core (clean folder, thumbnail, physics attribute/schema source layer). Use when a profile, validation report, or user request names FET_000_ISAAC; default to version `0.1.0`.

Mandatory acceptance policy for source-referenced CAD reconstruction; separates geometric fidelity from structural and persistence checks.

Inspect Kubernetes or OpenShift resources and perform only the mutations exposed by the chart-managed SRE API proxy.

Use when inspecting GPU/model capacity, live GPU utilization, GPU memory, power or temperature telemetry, model serving metrics such as queue depth, KV-cache usage or request latency, Prometheus/Thanos metric queries, model Routes or NodePorts, OpenShell…

Keep the memory inside its size limits by compacting rather than truncating, so it stays readable as it ages.

Check the memory against its schema and fix what can be fixed, so the store degrades visibly rather than silently.

Create and update the memory pages that the ranking job gates its top tier on, from evidence the selector collected.

Update a bounded preference policy from the user's repeated corrections, without ever overwriting the user's own choices.

Search the Outlook mailbox via Microsoft Graph to find and read emails that help answer user questions.

Look up and update single orders in an e-commerce backend. Use when the user asks about an order's status or wants one order changed.

Explains compiler and runtime errors. Use when the user pastes an error message and asks what it means.

Look up and update single orders in an e-commerce backend. Use when the user asks about an order's status or wants one order changed.

Look up and update single orders in an e-commerce backend. Use when the user asks about an order's status or wants one order changed.

Explains compiler and runtime errors. Use when the user pastes an error message and asks what it means.

Explains compiler and runtime errors. Use when the user pastes an error message and asks what it means.

A local greeting skill that never accesses the network

Documents manifest locations for an example skill.

Publish or update a NemoClaw pull request and follow its CI and automated reviews to completion.

Set up, repair, or check a NemoClaw contributor checkout. Also handles requested development CLI exposure, runtime onboarding, and pinned-agent launch.

Validate the staging NemoClaw Brev Launchable through its web journey or a deployed environment. Use for advisory validation outside automated E2E.

Triage, diagnose, debug, or fix failing or flaky NemoClaw E2E tests. Trace every assertion and downstream gate before a repair push or rerun. Excludes status-only and dispatch-only requests.

Classify one failed NemoClaw GitHub Actions job using bounded, redacted logs and optional retained artifacts.

Run or inspect NemoClaw live E2E evidence. Routes requested local execution, trusted GitHub dispatch, and release-evidence inspection.

Find the repository skill for a NemoClaw task or browse the skill catalog. Use when skill selection needs help.

Implement an accepted NemoClaw issue or repair a classified PR finding, with focused validation. Use for requested code or test changes.

Onboard a HuggingFace multimodal model (vision/audio/video + text) to the TensorRT-LLM PyTorch backend. Use when writing a new `tensorrt_llm/_torch/models/modeling_{vlm}.py` plus its input processor and weight mapper, or extending an existing VLM.

Runs model-level and module-level tests for TensorRT-LLM. First classifies the test scope (module test or model test), then dispatches to the appropriate workflow. Model tests are further classified by type (functionality/smoke test, benchmark, or…

Casebook of past successful and classic TensorRT-LLM optimizations (runtime/execution and kernel level) recorded as reusable decision precedents. Consult when deciding which optimization to apply for a classified bottleneck or a given config/model/hardware,…

Best practices for contributing code to TensorRT-LLM. Covers the official contribution process (issue tracking, fork workflow, DCO signing), coding guidelines, implementation workflow, common mistakes, testing strategy, commit hygiene, and review readiness.…

Launch and operate this repo's perf-optimize workflow, which iteratively APPLIES TensorRT-LLM serving optimizations — baseline benchmark at one concurrency or a Pareto curve of them (tok/s/user vs tok/s/gpu), analytical SOL projection on by default (via the…

Launch and operate this repo's perf-analyze workflow, which DIAGNOSES a TensorRT-LLM serving deployment without applying changes — benchmark at one concurrency or a Pareto curve of them (tok/s/user vs tok/s/gpu), analytical SOL projection on by default (via…

Design, implement, calibrate, and run deterministic TensorRT-LLM VisualGen L1 operation tests and L2 component-by-feature parity tests. Use for DiT, VAE, text-encoder, scheduler, parallelism, quantization, caching, compile, or model-specific component…

Systematic approach to exploring the TensorRT-LLM codebase before implementing new features or optimizations. Teaches how to discover existing infrastructure, trace code paths, and avoid reimplementing what already exists. Derived from real mistakes where…

Debug why an OpenShell gateway deployment is unhealthy, unreachable, or unable to create sandboxes. Use for gateway health failures, Docker/Podman runtime issues, Helm failures, Kubernetes scheduling, TLS or auth, gateway interceptors, supervisor middleware…

Guide agents through using the OpenShell CLI (openshell) for sandbox management, gateway registration, provider configuration and refresh, profile management, policy iteration, settings, service exposure, BYOC workflows, and attached-provider inference.…

Manually dispatch and iterate on the Release Canary workflow that smoke-tests published OpenShell artifacts (install.sh on macOS/Ubuntu/Fedora, Helm chart on kind) after each Release Dev publish. Use when changing `.github/workflows/release-canary.yml`,…

Start up, tear down, and configure the local Kubernetes development environment for OpenShell. Uses k3d (Docker-backed k3s) + Skaffold + Helm. Covers cluster lifecycle, optional add-ons (Keycloak OIDC, Envoy Gateway), HA testing, and port mappings. Trigger…

Generate sandbox security policies from plain-language requirements and optional REST API documentation. Produces L4 or fine-grained L7 network policies and ordered network middleware configuration. Use for API access rules, middleware host selection, failure…

Launch and supervise OpenShell gator agents. Use when starting gator on issues or PRs, checking gator sandboxes, building the gator sandbox image, restarting stuck gators, inspecting gator logs, or experimenting with gator harness/model overrides. Trigger…

Validate and monitor OpenShell GitHub issues and PRs using the gator:* state machine. Use when asked to triage issues/PRs for project validity, gate PRs, run gator, validate submissions, or monitor PRs toward merge readiness.

Watch and monitor GitHub Actions workflow runs using the gh CLI. Use when the user wants to check workflow status, watch a running workflow, view CI/CD jobs, or monitor build progress. Trigger keywords - watch pipeline, pipeline status, CI status, check…

Create a monthly NVIDIA Elements “What’s New” docs page, matching video, and concise, copy-ready summary from local release tags, tagged changelogs, package versions, and commit history. Use for a scheduled release PR, monthly update, release roundup,…

Create the 24–36 second marketing video for a monthly NVIDIA Elements “What’s New” post, rendered to projects/site/public/static/video/releases/MM-YYYY.webm from real Elements components, with feature snapshots inserted into the post. Use after the…

Verify every starter supported by the stable production NVIDIA Elements CLI from a fresh consumer environment, investigate failures, and prepare at most one focused repair pull request. Use for nightly or scheduled production starter verification and starter…

Use this skill by default for any UI-related work or with NVIDIA Elements (nve-*), including creating, editing, reviewing, or debugging HTML, CSS, layout, theming, components, applications, prototypes, Claude Artifacts, Codex Sites pages, and standalone UI…

Perform one small, scoped maintenance improvement to the Elements monorepo and open a single reviewable pull request. Use this skill for scheduled or unattended upkeep runs that improve unit test coverage for one file, fix one behavioral bug in one module, or…

Generate a production availability report for NVIDIA Elements packages, skill registry, and documentation.

Diagnose and improve browser WebGPU rendering performance, including frame pacing, draw and state-submission overhead, buffer uploads, shaders, and JavaScript work on the render critical path. Use for slow or stuttering GPU-backed interfaces, rendering…

Diagnose and resolve common issues including test failures, build errors, performance regressions, and development environment problems. Use this skill whenever the user reports something broken, failing, timing out, hanging, or producing unexpected results.…

Use this skill when integrating NVIDIA NeMo Fabric into a consumer application, service, evaluation harness, or platform through the typed Python SDK — translating the consumer's own application, job, or deployment config into an in-memory FabricConfig,…

Add or substantially change a first-party NVIDIA NeMo Fabric adapter, including repository packaging, discovery metadata, catalogs, CI wiring, and validation.

Verify whether a tagged NeMo Fabric release is deployed to crates.io, PyPI, and npm. Use when checking publication status for a specific release tag; not for publishing packages or creating tags.

Contribute documentation or example changes that stay aligned with NeMo Fabric public behavior

Create and push a signed, annotated NeMo Fabric beta tag from its release branch or from validated main. Use when cutting a beta tag; not for RC or stable release tags.

Create and push a signed, annotated NeMo Fabric release-candidate tag from its validated release branch. Use when cutting an RC tag; not for code freezes or stable release tags.

Create and push a signed, annotated NeMo Fabric stable release tag, then prepare unpublished GitHub Release and team announcement drafts for review. Use when cutting a stable release tag; not for beta or release-candidate tags.

Compare NVIDIA NeMo Fabric release refs and draft the authoritative GitHub Release body plus any warranted documentation-site release-note update. Use when preparing a stable release, creating patch-release notes, updating…

Perform a NeMo Relay project release-version change across project-owned manifests, dependencies, lockfiles, and generated attribution surfaces while keeping just set-version coverage complete. Do not use for dependency-only updates or adding a package when…

Verify whether a tagged NeMo Relay release reached GitHub Actions, tagged Go module source, crates.io, PyPI, and npm. Use for publication status of a specific tag; not for publishing packages or creating tags.

Create and push a signed, annotated NeMo Relay beta tag from its release branch or validated main. Use when cutting a beta tag; not for RC or stable tags.

Create and push a signed, annotated NeMo Relay release-candidate tag from its validated release branch. Use when cutting an RC tag; not for code freezes or stable tags.

Create and push a signed, annotated NeMo Relay stable tag, then prepare an unpublished GitHub Release draft for review. Use when cutting a stable release; not for beta or release-candidate tags.

Execute a requested NeMo Relay code freeze by cutting a release branch, updating nightly alpha configuration, bumping main, and preparing the required PR. Do not use for an ordinary version bump or release-note draft.

Add or change a public NeMo Relay runtime API across Rust and affected language bindings when no more specific maintainer skill owns the domain. Do not use for middleware additions, internal refactors, binding-local fixes, or docs-only changes.

Add a new NeMo Relay guardrail or intercept type, registration surface, or pipeline stage. Do not use for changing an existing middleware implementation without a new middleware contract.

Use when the user wants to run data augmentation on an annotated dataset, configure augmentation parameters, check augmentation status, or understand what each QA augmentation type does (BCQ, MCQ, GQA, DMCQ, DSQA, ENQA, WMCQ)

Use when running by-action VLM evaluation (per-action-clip inference + accuracy metrics) against the BP evaluation-ms HTTP API. Invoked as /sop-by-action-eval <inputs.yaml> [natural language parameter overrides]

Use when running the e2e evaluation pipeline (temporal segmentation + action recognition + accuracy) against the BP evaluation-ms HTTP API. Invoked as /sop-e2e-inference <inputs.yaml> [natural language parameter overrides]

Autonomous end-to-end orchestrator for SOP fine-tuning. Runs the full Import → Augment → DDM Train → VLM Train → Evaluate → RCA loop. Interprets RCA findings across DDM, VLM and augment axes, applies config fixes autonomously, and iterates until success…

Root cause analysis for SOP monitoring pipeline failures. Analyzes end-to-end evaluation logs, DDM temporal segmentation, VLM action recognition, training data, and fine-tuning configs to identify failure patterns and produce an evidence-driven RCA report…

Build a custom VSS SOP blueprint from the VSS 3.1 base, then deploy and test in a loop until fully operational. Use when asked to create the SOP blueprint structure, customize VSS compose for SOP, configure SOP services, set up the VSS agent for SOP, or…

Build the DS-SOP Docker image and deploy the VSS SOP blueprint end-to-end. Use when asked to deploy SOP, build DS SOP, install SOP, set up the SOP pipeline, verify SOP models, start the SOP blueprint, simulate RTSP for SOP, run the SOP API test, or tear down…

Use this skill when building, deploying, evaluating, debugging, or measuring latency for the DeepStream SOP Inference Microservice — a GPU-accelerated FastAPI service that detects whether operators perform assembly-line steps in order via event boundary…

Use when the user asks to "quantize a model", "run PTQ", "post-training quantization", "NVFP4 quantization", "FP8 quantization", "INT8 quantization", "INT4 AWQ", "quantize LLM", "quantize MoE", "quantize VLM", or needs to produce a quantized HuggingFace…

Evaluates accuracy of quantized or unquantized LLMs using NeMo Evaluator Launcher (NEL). Triggers on "evaluate model", "benchmark accuracy", "run MMLU", "evaluate quantized model", "run nel". Handles deployment, config generation, and evaluation execution.…

Audit merged bug-fix PRs and release NVBugs for missing cherry-pick labels, then cherry-pick labeled PRs into a release branch and open a PR. Use when asked to "cherry-pick PRs for release/X.Y.Z", "pick PRs to release branch", "verify cherry-pick labels", or…

Deterministic end-to-end driver for day-0 quantized-checkpoint releases — chains PTQ → evaluation → comparison with enforced gates between stages (the evaluation stage deploys the checkpoint itself), and returns a publish decision (ACCEPT / REGRESSION /…

Run the AgentX agentic serving benchmark with the SemiAnalysis harness. Use for AgentX setup, dataset downloads, concurrency sweeps, or serving latency and throughput comparisons.

Establish baseline-vs-candidate evaluation plans, delegate missing evaluations, compare validated results, and decide quantization feasibility. Use when the user asks to compare baseline vs quantized runs, explain an accuracy drop/regression, verify whether a…

Use when the user asks to find, search for, or optimize the best quantization recipe for a model, including direct requests like "find the best quantization recipe and generate a PTQ checkpoint." Guides the multi-candidate loop: choose compute-vs-memory…

Train, debug, and validate a speculative decoding draft model (EAGLE3, DFlash, DSpark, Domino) through the ModelOpt launcher pipeline. Use when the user wants to add a new model to a draft-training pipeline, asks why a pipeline run failed, wants experiment…