What does this skill do, and when should you use it?
This skill handles network discovery, download, and staging of NVIDIA Jetson Linux BSP artifacts. It filters the official Jetson Linux archive for a concrete release that explicitly supports the active target platform, then obtains the BSP, sample rootfs, public sources, cross-toolchain, and user documentation. Files are placed where the downstream Jetson setup skills expect them, with checks for disk space, filenames, and archive formats. The skill is download-only: extraction, source initialization, document registration, and target-profile edits belong to other skills.
Reads the active target profile and optional prefilled BSP release; accesses NVIDIA’s Jetson Linux archive and linked release pages; discovers and downloads the BSP tarball, sample rootfs tarball, public_sources.tbz2, x-tools.tbz2, Developer Guide, and User Guide; places image archives in the Image or configured BSP root, sources and toolchain in Downloads, and guides in Documents; checks available disk space, reuses non-empty existing files, downloads through temporary .part files before renaming, and validates release tokens and bzip2 formats for key archives.
- A Jetson developer has selected Auto Setup and needs the required BSP artifacts before initialization skills run.
- A platform engineer needs an offline-ready workspace containing the Jetson image archives, sources, and cross-toolchain for an active target.
- An integration engineer needs to select a concrete Jetson Linux release explicitly supported by the current reference devkit.
- A maintainer wants existing downloads reused safely without overwriting destination files.
How do you install this skill?
- Downloaded artifacts are not required to undergo cryptographic integrity verification; confirm official checksums or signatures before use.
- The retry procedure conflicts with the stated no-retry limitation; the implementation should resolve this discrepancy.
- Core operation depends on developer.nvidia.com and possible browser/EULA flows; confirm mainland-China connectivity and authentication availability.
- The skill card states CC-BY-4.0 AND Apache 2.0, while SKILL.md states only Apache-2.0; clarify the applicable license.
- Shell / CLI
- Network access
- Local filesystem
curlfiledfnpx skills CLI
Install the skill from NVIDIA/skills with the skills CLI:
npx skills add nvidia/skills --skill jetson-download-bsp --yes
As documented in the README, you can also add --agent codex, --agent claude-code, --agent cursor, or --agent kiro-cli.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/NVIDIA/skills.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/skills/jetson-download-bsp" ~/.claude/skills/
rm -rf "$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?
Resolve an active target profile first, then use /jetson-quick-start and select Auto Setup, or ask the agent to download the NVIDIA BSP artifacts required by the current workspace. Choose a concrete release explicitly supported by the active platform, such as R36.4.4, 38.2, or 38.2.1; family placeholders such as R38.x are not accepted. After downloading, hand off to /jetson-init-image, /jetson-init-source, and /jetson-link-docs.
What are this skill's strengths and limitations?
- Covers the BSP, sample rootfs, public sources, cross-toolchain, and both named guides.
- Requires a concrete supported release and does not silently substitute another version.
- Discovers artifacts through official page links instead of constructing NVIDIA download URLs.
- Supports reuse of existing files, disk-space checks, failure reporting, and key artifact validation.
- Requires access to developer.nvidia.com and the NVIDIA archive pages.
- Requires an active target profile and the expected workspace layout.
- Does not extract archives, run apply_binaries.sh, mount sources, register documents, or edit the active profile.
- The source provides no platform matrix, test suite, or success-rate evidence; browser/EULA downloads may require manual user handling.
- The SKILL.md says to retry discovery three times but later lists “No retry / mirror logic” as a limitation, so retry behavior is inconsistently documented.
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 |
|---|---|---|---|---|
| Jetson BSP Download Assistant this page ✓ NVIDIA · Official | 51 · Use with care | ★ 3.8k | today | Apache-2.0 |
| Jetson nvpmodel Power Mode Customizer ✓ NVIDIA · Official | 54 · Use with care | ★ 3.8k | today | Apache-2.0 |
| Jetson BSP Image Initializer ✓ NVIDIA · Official | 52 · Use with care | ★ 3.8k | today | Apache-2.0 |
| Jetson Device Info Snapshot ✓ NVIDIA · Official | 58 · Recommended | ★ 3.8k | today | Apache-2.0 |
| Jetson Health Snapshot ✓ NVIDIA · Official | 56 · Use with care | ★ 3.8k | today | Apache-2.0 |
How did FollowSkills review this skill?
The skill clearly limits itself to downloading, declares public data, requires a concrete user-selected release, and addresses overwrite confirmation, EULA handling, disk checks, failure reporting, and temporary .part files before rename. It does not require checksums, signatures, or supply-chain integrity verification, and it performs external network and workspace writes, so 7 points are deducted.
The workflow specifies paths, retries, disk checks, file-type validation, and controlled failure feedback. However, the three-attempt WebFetch procedure conflicts with the stated limitation of having no retry or mirror logic, and there is no executable implementation or committed test coverage for key paths. The static score is therefore limited to 7.
Invocation conditions, inputs, outputs, non-goals, and release-selection rules are well defined. Chinese-language interaction and mainland-China reachability are not addressed, and the workflow assumes a particular workspace and access to developer.nvidia.com, so 6 points are deducted.
The skill includes versioning, a license field, references, prerequisites, limitations, troubleshooting, and handoff documentation. It lacks the recommended Instructions and Examples sections, has an author-format issue noted by the benchmark, provides no changelog or clearly assigned maintenance path, and its license wording is not fully consistent with the skill card and repository license. Seven points are deducted.
The objective and destination paths are concrete, and the supplied two-task benchmark reports 88% Codex correctness and 66% effectiveness. The sample is small and positive-only; actual downloads, link discovery, and archive compatibility were not independently verified in this static review, and the result depends on external network access and user intervention. Nine points are deducted, yielding 6.
The supplied material includes revision-pinned source context, two evaluation tasks, and reported benchmark results, but no committed tests or CI reproducing this skill's key paths and no independent corroborating sources. Because this is a static review with no execution, the score is 3.
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 →