ncu-profile-analysis

ncu-profile-analysis is a skill for Claude Code from lucifer1004/VeloQ. It costs 61 tokens per session (3,222 once invoked), scanned C, original, MIT.

Instructions for analyzing Nsight Compute reports, which record detailed GPU kernel performance data. They use VeloQ, a command-line report reader, to examine occupancy, memory use, instruction activity, stalls, and source-code correlation.

In plain words
What is it for?
Use them to inspect .ncu-rep or compressed .ncu-repz files, investigate kernel bottlenecks, connect metrics to source lines, and export findings as tables or CSV.
Why use it?
They provide a consistent way to find why an individual GPU kernel is slow without manually parsing report files or hidden generated data.

Skill for Claude Code

Written for Claude Code: shipped in a Claude Code plugin.

Part of the VeloQ plugin — 3 skills shipped together

Good fit Use them to inspect .ncu-rep or compressed .ncu-repz files, investigate kernel bottlenecks, connect metrics to source lines, and export findings as tables or CSV.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/lucifer1004/veloq/ncu-profile-analysis
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add lucifer1004/VeloQ --skill ncu-profile-analysis
Clone the repo
git clone --depth 1 https://github.com/lucifer1004/VeloQ

Made for: Claude Code.

Or install VeloQ, the plugin that ships this one along with the rest of its 3 skills.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for ncu-profile-analysis

README.md
[![agentmods](https://agentmods.dev/badge/skills/lucifer1004/veloq/ncu-profile-analysis/github.svg)](https://agentmods.dev/skills/lucifer1004/veloq/ncu-profile-analysis)
Your own site
<a href="https://agentmods.dev/skills/lucifer1004/veloq/ncu-profile-analysis"><img src="https://agentmods.dev/badge/skills/lucifer1004/veloq/ncu-profile-analysis/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for ncu-profile-analysis

Your own site · 80×15
<a href="https://agentmods.dev/skills/lucifer1004/veloq/ncu-profile-analysis"><img src="https://agentmods.dev/badge/skills/lucifer1004/veloq/ncu-profile-analysis.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 61 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 3,222 The whole file, excluding the scripts and references it only reads on demand.
Security scan C 2 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00061 $0.03222
Opus 5 $0.00030 $0.01611
Sonnet 5 $0.00012 $0.00644
Haiku 4.5 $0.00006 $0.00322

Measured 11d ago against content hash 0e8fa886e3ae, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade C, and why

ncu-profile-analysis scanned grade C with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Downloads and executes remote codehighSupply chain

curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.

curl -fsSL https://raw.githubusercontent.com/lucifer1004/veloq/main/scripts/install.sh | bash

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

curl -fsSL https://raw.githubusercontent.com/lucifer1004/veloq/main/scripts/install.sh | bash
plugins/veloq/skills/ncu-profile-analysis/SKILL.md · 250 lines

How it starts

The opening of the file, as written. The whole thing — 250 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Nsight Compute Profile Analysis

Use this skill for Nsight Compute .ncu-rep and zstd-compressed .ncu-repz kernel reports and kernel-internal bottleneck analysis. NCU explains why a kernel is slow internally; NSys explains what ran when. Use veloq ncu as the evidence extractor for report data, then decide whether the evidence points to occupancy, memory, scheduler, instruction, launch, or source-line work.

This skill requires the VeloQ CLI on PATH. If veloq is missing, install it before analysis.

Tool Boundary

Use veloq ncu verbs as the analysis interface. Do not parse <file>.veloq/ncu-native.json.gz, disasm cache files, or other .veloq/ artifacts directly unless the user explicitly asks for raw artifact inspection or you are developing VeloQ itself.

Detection

veloq info path/to/report.ncu-rep
veloq ncu summary --help

If veloq is missing, install it:

# Pre-built binary (Linux x86_64/aarch64, macOS x86_64/arm64)
curl -fsSL https://raw.githubusercontent.com/lucifer1004/veloq/main/scripts/install.sh | bash

# Or build from source — clone the repo first, then:
cargo build --release -p veloq

Envelope contract

JSON output is the agent contract. Success:

{
  "schema": "v1",
  "source": { "kind": "ncu", "version": "v1" },
  "command": "ncu.launches",
  "trace": { "kind": "ncu", "path": "..." },
  "data": {
    "count": 50,
    "total_matched": 1234,
    "rows": [
      /* ... */
    ],
    "auxiliary": {
      /* per-verb */
    }
  }
}

trace_span is omitted on NCU verbs (NCU reports do not carry the NSys-style primary-execution window). Every list verb returns canonical data.rows[] with a stable per-row key:

  • ncu summary → one totals row (key: "totals")
  • ncu launches / inspectkey = "launch:<idx>"
  • ncu disasmkey = "kernel|<function_name>"; requested launch id is echoed under data.auxiliary.row_id
  • ncu metrics (long form) → key = "launch:<idx>|counter:<name>"
  • ncu metrics --per-launchkey = "launch:<idx>"
  • ncu source-metrics --by linekey = "launch:<idx>|line:<file>:<line>"
  • ncu source-metrics --by sasskey = "launch:<idx>|sass:0x<addr>"
  • ncu source-metrics --by filekey = "launch:<idx>|file:<file>"
  • ncu warp-stalls --by linekey = "launch:<idx>|line:<file>:<line>"
  • ncu warp-stalls --by sasskey = "launch:<idx>|sass:0x<addr>"
  • ncu warp-stalls --by reasonkey = "launch:<idx>|reason:<stall_reason>"
  • ncu sourceskey = "source:<idx>"
  • ncu ranges / graphskey = "<kind>:<idx>"

Read the full file on GitHub · 250 lines

Files

What ships with it

7 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 11d ago First seen · 250 lines · 61 tokens per session scan C 0e8fa886e3ae

Subscribe to this mod's changes

ncu-profile-analysis is a skill published in the GitHub repository lucifer1004/VeloQ (120 stars, last pushed 20d ago), licensed MIT. It adds 61 tokens to every session and 3,222 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it C with 2 findings (downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other skills, from other repositories

ncu-cuda-profiling

Automated NCU (Nsight Compute) profiling workflow with full metrics collection and persistent storage.

maxiaosong1124/ncu-cuda-profiling-skill · 26 tokens

cuda-kernel-optimizer

Use when optimizing or diagnosing GPU workload performance with CUDA, CUTLASS, Triton, PyTorch, vLLM or TensorRT-LLM; interpreting NCU, Nsys or PyTorch Profiler evidence; or preparing the workload and correctness checks for such optimization. General coding, skill maintenance and conceptual GPU questions do not…

troycheng/cuda-kernel-optimizer · 77 tokens

ncu-cuda-profiling

Automated NCU (Nsight Compute) profiling workflow with full metrics collection and persistent storage.

wenyi-li/awesome-agent-kernel-skills · 26 tokens

modal-serverless-gpu

Run approved CPU or GPU work through OpenScience computejob on the user's configured Modal account. Use for isolated scientific scripts, dependency provisioning, durable outputs, logs, status, cancellation, and recovery. Never invoke the Modal SDK or CLI directly.

synthetic-sciences/openscience · 54 tokens

cuopt-skill-evolution

After solving a non-trivial problem, detect generalizable learnings and propose skill updates. Always active — applies to every interaction.

NVIDIA/cuopt · 33 tokens

profiling

Diagnose frame-rate bottlenecks (CPU vs GPU bound FIRST), control Unreal Insights traces, sample live frame times, and annotate performance captures. Use when FPS is low/bad, the game is slow, or you need to find what is limiting the frame rate.

kevinpbuckley/VibeUE · 57 tokens