An index of instructions for HBDK4, a toolchain that converts machine-learning models into files runnable on Horizon hardware. It covers importing, serializing, quantizing, modifying, compiling, inspecting, and validating models.

Use it when exporting ONNX models, loading or saving BC files, converting models to a target hardware version, compiling HBM files, editing model inputs or outputs, or validating inference.

It provides a task-based map of the model-conversion workflow so the appropriate instructions can be selected for each stage.

skills — Skill for Claude Code and Codex by HorizonRobotics

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 HorizonRobotics/OE-Skills --skill hbdk-manual

Clone the repo

git clone --depth 1 https://github.com/HorizonRobotics/OE-Skills

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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/horizonrobotics/oe-skills/hbdk-manual/github.svg)](https://agentmods.dev/skills/horizonrobotics/oe-skills/hbdk-manual)

Your own site

<a href="https://agentmods.dev/skills/horizonrobotics/oe-skills/hbdk-manual"><img src="https://agentmods.dev/badge/skills/horizonrobotics/oe-skills/hbdk-manual/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.

Your own site · 80×15

<a href="https://agentmods.dev/skills/horizonrobotics/oe-skills/hbdk-manual"><img src="https://agentmods.dev/badge/skills/horizonrobotics/oe-skills/hbdk-manual.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 21 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 1,107 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.

Model	Per session	Once invoked
Fable 5.1	$0.00021	$0.01107
Opus 5	$0.00010	$0.00553
Sonnet 5	$0.00004	$0.00221
Haiku 4.5	$0.00002	$0.00111

Measured 12d ago against content hash 604e77e2737e, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

skills 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 12d 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.

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.

horizon/skills/hbdk/hbdk-manual/SKILL.md · 89 lines

How it starts

The opening of the file, as written. The whole thing — 89 lines — stays where its author put it; the contents beside it link to each section on GitHub.

基于HBDK4版本 4.9.7 编写，Docker镜像版本 v3.9.0_rc4。

流程总览

.bc模型有两个来源：

PTQ流程：ONNX模型通过 hbdk4.compiler.onnx.export 导出（ptq.onnx → ptq.bc）
QAT流程：由其它模块从 torch.nn.Module 导出（qat.bc），该torch模型不在hbdk工作流内，由外部提供

PTQ: ONNX模型(ptq.onnx) → 导出HBIR(ptq.bc) ─┐
                                              ├→ 定点化(quantized.bc) → 编译(HBM) → 部署
QAT: torch.nn.Module → qat.bc（外部提供） ────┘
                                                    │              │
                                               插入预处理节点    性能评估
                                               查询/修改模型     推理验证
                                                                    │
                          完整工作流参考 → references/workflow_reference.md

Skill列表

模型导入与序列化

Skill	说明
model-export.md	将ONNX模型导出为伪量化HBIR模型（.bc文件），支持模型统计和可视化
model-serialization.md	将HBIR模型保存到.bc文件，或从.bc文件加载HBIR模型

模型转换与编译

Skill	说明
model-convert.md	将伪量化HBIR模型转换为指定march的定点模型
model-compile.md	将定点化HBIR模型编译为HBM文件

模型修改

Skill	说明
model-info-query.md	查询和修改HBIR模型的输入输出、march信息、tensor名称、精度配置等
insert-nodes.md	在HBIR模型输入输出参数上插入预处理节点（pyramid/preprocess/resize/transpose/split/rle）
remove-io-nodes.md	删除HBIR模型输入输出相邻节点或指定类型/名称的节点

HBM模型

Skill	说明
model-inference.md	推理HBIR和HBM模型，包括本地推理和远程BPU推理
hbm-perf.md	对HBM模型进行性能分析，包括静态perf和动态perf
hbm-modify.md	修改HBM模型描述信息和名称，查询量化信息

参考

文档	说明
references/api_reference.md	API快速参考
references/workflow_reference.md	端到端工作流参考

快速入门路径

场景：我有一个ONNX模型，想部署到BPU

model-export.md — 导出ONNX为HBIR
model-convert.md — 定点化转换
model-compile.md — 编译为HBM
hbm-perf.md — 评估性能

Read the full file on GitHub · 89 lines

Files

What ships with it

13 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.

12d ago First seen · 89 lines · 21 tokens per session scan A 604e77e2737e

Subscribe to this mod's changes

skills is a skill published in the GitHub repository HorizonRobotics/OE-Skills (19 stars, last pushed 2mo ago), licensed Apache-2.0. It adds 21 tokens to every session and 1,107 once invoked, about $0.0001 per session on Opus 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 2026-08-30.

Data and AI Hardware and IoT

spark-environment-setup

Set up a working ML training/inference environment on NVIDIA DGX Spark (GB10, aarch64, CUDA 13). Use when installing PyTorch/Unsloth/TRL/vLLM on DGX Spark, hitting libcudart or wheel-ABI errors on aarch64, or choosing between NGC containers and bare pip installs.

wshobson/agents · 76 tokens

spark-memory-thermal-ops

Manage unified memory and thermals during long-running ML jobs on NVIDIA DGX Spark. Use when planning memory headroom for a training run on GB10, when a job OOMs on unified memory, or when monitoring temperature and power during multi-hour training.

wshobson/agents · 59 tokens

spark-training-gotchas

Preflight and diagnose the ten known failure modes for ML training on NVIDIA DGX Spark. Use when a training run on DGX Spark fails to start, OOMs below the 128GB limit, slows down mid-run, or before any multi-hour training job on GB10.

wshobson/agents · 63 tokens

llama-cpp

Runs LLM inference on CPU, Apple Silicon, and consumer GPUs without NVIDIA hardware. Use for edge deployment, M1/M2/M3 Macs, AMD/Intel GPUs, or when CUDA is unavailable. Supports GGUF quantization (1.5-8 bit) for reduced memory and 4-10× speedup vs PyTorch on CPU.

davila7/claude-code-templates · 76 tokens

minicpm5-deploy-vllm-ascend

Deploy MiniCPM5-2B with vLLM on Huawei Ascend NPU using vLLM-Ascend. Use when the user mentions vLLM-Ascend, Ascend NPU, Huawei Ascend, CANN, torchnpu, davinci devices, or wants an OpenAI-compatible MiniCPM5 server on Ascend hardware.

OpenBMB/MiniCPM · 87 tokens

minicpm5-deploy-litert

Run MiniCPM5-2B or MiniCPM5-1B on-device with Google's LiteRT-LM runtime — the litert-lm CLI or its OpenAI-compatible server on a desktop, the Kotlin API or the AI Edge Gallery app on Android, the same .litertlm bundle on CPU or GPU. Use when the user says "LiteRT", "LiteRT-LM", "litertlm", ".litertlm", "Android"…