nvidia-skill-finder
Use for NVIDIA-related requests where an NVIDIA skill might help, even if the user did not ask for a skill. Trigger on NVIDIA products, hardware, software,…
Use when you need to print Jetson BSP info (L4T version, board configs, rootfs state) from a Linux_for_Tegra root on the host PC. This is an example skill.
$ npx -y skills add NVIDIA/skills --skill jetson-print-bsp-info --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/jetson-print-bsp-infoContext preview
The summary Claude sees to decide when to auto-load this skill.
Use when you need to print Jetson BSP info (L4T version, board configs, rootfs state) from a Linux_for_Tegra root on the host PC. This is an example skill.
name: jetson-print-bsp-info description: Use when you need to print Jetson BSP info (L4T version, board configs, rootfs state) from a Linux_for_Tegra root on the host PC. This is an example skill. version: 0.0.1 license: "Apache-2.0" metadata: author: "Jetson Team" tags: [jetson, bsp, info] languages: [bash] data-classification: public
Prints a concise summary of a Jetson Linux_for_Tegra (BSP) tree on the host PC.
This skill is intended as a reference example for the `jetson-bsp-skills` repo and the NVIDIA-wide skills CI. It performs read-only inspection — no flashing, no rootfs changes.
Capture a baseline snapshot of a Linux_for_Tegra BSP tree (release, board configs, rootfs state) before flashing, so issues like "wrong L4T version" or "rootfs never populated" are caught early.
Run each step in order and print the captured values into the report shown under [Output format](#output-format).
1. **Resolve** `L4T_ROOT` and **validate** the directory is a Linux_for_Tegra root — exit early otherwise. `flash.sh` and `nv_tegra/` are the two anchor artifacts that every BSP ships:
L4T_ROOT="${L4T_ROOT:-$PWD}"
if [ ! -f "$L4T_ROOT/flash.sh" ] || [ ! -d "$L4T_ROOT/nv_tegra" ]; then
echo "Not a Linux_for_Tegra root: '$L4T_ROOT' (missing flash.sh or nv_tegra/)"
exit 1
fi
echo "$L4T_ROOT"2. **Extract** the L4T release header line. The canonical host-side location is `nv_tegra/nv_tegra_release`; the same file is copied into the rootfs by `apply_binaries.sh`. Only the first line is useful — the rest is a long list of library SHAs:
head -1 "$L4T_ROOT/nv_tegra/nv_tegra_release" 2>/dev/null \
|| head -1 "$L4T_ROOT/rootfs/etc/nv_tegra_release" 2>/dev/null \
|| echo "L4T release info not found"3. **List** supported board config files and **join** them onto one comma-separated line:
(cd "$L4T_ROOT" && ls *.conf 2>/dev/null) | paste -sd, -
4. **Check** whether the rootfs has been populated. An empty `rootfs/` means `apply_binaries.sh` has not been run yet:
if [ -f "$L4T_ROOT/rootfs/etc/passwd" ]; then
echo "populated"
else
echo "empty"
fiPrint a short report with these sections, one line each where possible:
L4T root: <path> L4T release: <release header line> Board configs: <comma-separated list> Rootfs: populated | empty
Example output on an Orin AGX BSP (L4T R36):
L4T root: $HOME/Linux_for_Tegra L4T release: # R36 (release), REVISION: 3.0 Board configs: jetson-agx-orin-devkit.conf,jetson-orin-nano-devkit.conf Rootfs: populated
Example output on a freshly untarred BSP where `apply_binaries.sh` has not been run yet:
L4T root: /tmp/Linux_for_Tegra L4T release: # R39 (release), REVISION: 0.0 Board configs: jetson-agx-thor-devkit.conf Rootfs: empty
Each command falls back to a clearly labeled `"... not found"` string if the underlying file is missing — the skill never errors out mid-report. If `L4T_ROOT` does not contain `flash.sh` and `nv_tegra/`, exit early with a clear "not a Linux_for_Tegra root" message rather than printing misleading info.
**Cause:** `L4T_ROOT` points at a parent directory, an extracted rootfs, or an unrelated path. **Solution:** Point `L4T_ROOT` at the directory that contains `flash.sh` (typically `Linux_for_Tegra/`).
**Cause:** Neither `nv_tegra/nv_tegra_release` nor `rootfs/etc/nv_tegra_release` exists — the BSP tarball may be incomplete or `apply_binaries.sh` was never run. **Solution:** Re-extract the BSP tarball or run `apply_binaries.sh` to populate the rootfs.
Official, NVIDIA-verified Agent Skills for Claude Code, Codex, and other coding agents.
Use for NVIDIA-related requests where an NVIDIA skill might help, even if the user did not ask for a skill. Trigger on NVIDIA products, hardware, software,…
Official NVIDIA-authored guidance for NVIDIA cuDF GPU DataFrames, pandas acceleration, dask-cuDF, ETL, joins, groupby, CSV/Parquet I/O, nullable semantics, and…
Use when asked to install, deploy, run, validate, troubleshoot, or stop NVIDIA AI-Q Blueprint infrastructure.
Use when asked to run deep research or AI-Q research through a reachable NVIDIA AI-Q Blueprint backend.
Calibrate a new dataset from live RTSP camera streams via the AutoMagicCalib REST API. Use when the user provides RTSP URLs or asks to calibrate live cameras;…
Run end-to-end calibration on the shipped sample dataset (sdg_08_2_sample_data_010926.zip) against a running AMC microservice. Use when user says 'test sample…