agentmods.dev

rdk-npu-inference

rdk-npu-inference is a skill for Claude Code, Codex from D-Robotics/moss. It costs 48 tokens per session (627 once invoked), scanned A, original, MIT.

A command-line workflow for running and measuring neural-network models on RDK X5 and S100 boards. It uses the board's NPU tools to load models, run inference, and report latency and frame rate.

In plain words
What is it for?
Use it to inspect a model, benchmark its performance, or run inference on a binary input over SSH.
Why use it?
It gives you repeatable checks that a model loads on the device and evidence of how quickly it runs there.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

not rated 143repo yesterday A scan 48 tokens original MIT

Good fit Use it to inspect a model, benchmark its performance, or run inference on a binary input over SSH.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/d-robotics/moss/rdk-npu-inference
View source ↗ D-Robotics/moss
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 D-Robotics/moss --skill rdk-npu-inference
Clone the repo
git clone --depth 1 https://github.com/D-Robotics/moss

Made for: Claude Code, Codex.

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 rdk-npu-inference

README.md
[![agentmods](https://agentmods.dev/badge/skills/d-robotics/moss/rdk-npu-inference/github.svg)](https://agentmods.dev/skills/d-robotics/moss/rdk-npu-inference)
Your own site
<a href="https://agentmods.dev/skills/d-robotics/moss/rdk-npu-inference"><img src="https://agentmods.dev/badge/skills/d-robotics/moss/rdk-npu-inference/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 rdk-npu-inference

Your own site · 80×15
<a href="https://agentmods.dev/skills/d-robotics/moss/rdk-npu-inference"><img src="https://agentmods.dev/badge/skills/d-robotics/moss/rdk-npu-inference.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 48 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 627 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 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 $0.00048 $0.00627
Opus 5.5 $0.00019 $0.00251
Sonnet 5.5 $0.00010 $0.00125
Haiku 4.5 $0.00005 $0.00063

Measured yesterday against content hash 942400d38c0f, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

rdk-npu-inference scanned grade A with 0 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 yesterday.

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.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.moss/skills/rdk-npu-inference/SKILL.md · 54 lines

What it actually says

RDK NPU inference via device tools

Load and verify a model

device_exec: hrt_model_exec model_info --model_file /path/to/model.bin

Success shows Load model to DDR cost Xms plus per-model info. The model builder version in the output identifies the toolchain that compiled it — a mismatch with the running runtime means a version-skew, not a runtime failure.

Benchmark latency and FPS

device_exec: hrt_model_exec perf --model_file /path/to/model.bin --thread_num 2

Runs 200 frames, then prints Average latency is: X ms and Frame rate is: Y FPS. Record both as evidence.

Run inference on an input

device_exec: hrt_model_exec infer --model_file /path/to/model.bin --input_bin_path /path/to/input.bin

Board-shipped test models

Board path What it is
/app/multimedia_samples/chip_base_test/.../HBDK3_MODEL_2K/x5/*.hbm BPU self-test (2K resolution)
/app/multimedia_samples/sunrise_camera/.../model_zoom/mobilenetv2_224x224_nv12.bin MobileNetV2 classification
/app/multimedia_samples/sunrise_camera/.../model_zoom/yolov5s_672x672_nv12.bin YOLOv5s detection
/opt/hobot/model/x5/basic/*.bin TROS node models

Real baselines (RDK X5, 2026-11)

Model Latency FPS Source
MobileNetV1 224x224 (OE-compiled) 1.58 ms 1247 cloud→board deploy
MobileNetV2 224x224 (preinstalled) 1.6-1.8 ms 520-600 sunrise_camera
HBDK3 2K test model 44 ms 45 chip_base_test

Rules

  • A golden result ERROR after bpu0 test sucess in the chip_base_test means a missing reference file, not a BPU failure.
  • One heavy NPU process at a time; GPU/NPU contention yields intermittent depth/conversion failures.
  • Background a long NPU run safely: nohup CMD > log 2>&1 < /dev/null & echo $! > pid.
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. yesterday First seen · 54 lines · 48 tokens per session scan A 942400d38c0f

Subscribe to this mod's changes

rdk-npu-inference is a skill published in the GitHub repository D-Robotics/moss (143 stars, last pushed yesterday), licensed MIT. It adds 48 tokens to every session and 627 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-11-08.

Data and AIHardware and IoT
Related

rdk-npu-inference compared with skills from other repositories

rdk-npu-inference next to 6 skills from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
rdk-npu-inferenceD-Robotics/mossSkill143yesterdayMITA48Claude Code, Codex
spark-environment-setupwshobson/agentsSkill42kyesterdayMITA76Claude Code
spark-memory-thermal-opswshobson/agentsSkill42kyesterdayMITA59Claude Code
spark-training-gotchaswshobson/agentsSkill42kyesterdayMITA63Claude Code
inference-serving-llama-cppdavila7/claude-code-templatesSkill33ktodayMITA80Claude Code, Codex
amc-run-rtsp-calibrationNVIDIA/skillsSkill3.9k2d agoApache-2.0A59Claude Code, Codex
deepstream-generate-pipelineNVIDIA/skillsSkill3.9k2d agoApache-2.0A108Claude Code