DESIGN.md@skills/jetson-memory-audit/referencesA
DESIGN.md@skills/jetson-memory-audit/references from nvidia/skills is a DESIGN.md file: the design rules coding agents follow in that repo.
mdr add nvidia/skills/skills/jetson-memory-audit/references/DESIGN.md@git:20260622.87053a0Pins git:20260622.87053a0 in mdr.lock. Pin the fingerprint instead.
- Audit grade
- A, 14 of 14 checks passed
- Kind
- DESIGN.md file
- Repository
- nvidia/skills
- Path
skills/jetson-memory-audit/references/DESIGN.md- Versions recorded
- 1
- Latest version
- git:20260622.87053a0, committed Jun 22, 2026
- Size
- 2,628 bytes
- Repository stars
- 3,871
- License
- Apache-2.0
Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
# Design notes — jetson-memory-audit
## Why audit.sh delegates to snapshot.sh
`audit.sh` was originally written as a standalone memory snapshot. It duplicated
logic from `jetson-diagnostic/scripts/snapshot.sh` and diverged in two important
ways that caused it to produce wrong or incomplete results on some Jetsons.
### 1. No nvidia-smi capability probe (Thor and later)
`audit.sh` always read NvMap from `/sys/kernel/debug/nvmap/iovmm/clients` as the
GPU memory source. On Jetsons using the unified `nvidia.ko` driver (Thor T5000/T4000),
NvMap debugfs is present but empty — the authoritative source is
`nvidia-smi --query-compute-apps`. Without the capability probe, `audit.sh`
silently reported 0 KB of GPU memory on Thor, giving a misleading picture of
what was actually holding memory.
`snapshot.sh` introduced a probe — `nvsmi_useful()` — that distinguishes the two
driver stacks at runtime:
```bash
nvsmi_useful() {
command -v nvidia-smi >/dev/null 2>&1 || return 1
local name
name=$(nvidia-smi --query-gpu=name --format=csv,noheader 2>/dev/null | head -n1)
case "$name" in
''|*'(nvgpu)'*) return 1 ;; # nvgpu stub — use NvMap
…Read the whole file at its exact version.
How to install
mdr add nvidia/skills/skills/jetson-memory-audit/references/DESIGN.md@sha256:55129b398c29c8cdThe line at the top of this page pins the label git:20260622.87053a0, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
[](https://markdownregistry.com/a/art_bfd4qff7cwbqay5e)
3 badge image fetches in 30 days. What the badge shows.
Versions
The registry has recorded 1 version of this file, committed on Jun 22, 2026.
Audit of the latest version
- pass: Size between 200 bytes and 200 KB (2628 bytes)
- pass: No zero-width or bidi control characters
- pass: No instruction hidden inside an HTML comment
- pass: No link to an exfiltration or paste host
- pass: No credential-shaped string
- pass: No instruction to send local credentials anywhere
- pass: No text hidden with inline styles
- pass: No prompt-injection phrasing
- pass: No curl or wget piped into a shell
- pass: No recursive delete of root, home or parent
- pass: No instruction to read or print local credentials
- pass: No base64 blob over 200 characters
- pass: No link to a raw IP address
- pass: No script tag
Source
nvidia/skills · 3,871 stars · license Apache-2.0 · pushed 2026-11-05 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_bfd4qff7cwbqay5e GET https://markdownregistry.com/api/v1/resolve?ref=nvidia/skills/skills/jetson-memory-audit/references/DESIGN.md GET https://markdownregistry.com/api/v1/blob/55129b398c29c8cd943f09675b977dfc43dc43d5ae363b1dd50ec404e5a2df24
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
Guides: what DESIGN.md is and how agents use it · which agent file does what.