Scanned skills

nvidia/skills deepstream-import-vision-model

Checked the catalogued skill at deepstream-import-vision-model.

SkillTrust corpus scan 3 October 2026 v0.13.0 via skillsh Verify on GitHub
F AI-triaged 98
F Security

F; retained sample includes a high finding · SD-007 · what this rule accepts as a false positive →

D Permissions

D; retained sample includes a high finding · SD-003 · what this rule accepts as a false positive →

B Transparency

B; retained sample includes a medium finding · SD-007 · what this rule accepts as a false positive →

showing 10 of 98
transparencymedium SD-007
evals/evals.json
transparencymedium SD-007
evals/evals.json
transparencymedium SD-007
evals/evals.json
transparencymedium SD-007
evals/evals.json
transparencymedium SD-007
evals/evals.json
transparencymedium SD-007
evals/evals.json
permission_hygienehigh SD-003
install.ps1
securityhigh SD-007
install.ps1
securityhigh SD-007
install.ps1
securityhigh SD-007
install.sh
5 suppressed by AI triage
permission_hygienehigh SD-003 benign_example
SKILL.md

The path is part of documentation, not executed, so no traversal risk.

permission_hygienehigh SD-003 benign_example
SKILL.md

The absolute path is only referenced in comments; no code uses it to access filesystem.

permission_hygienehigh SD-003 benign_example
references/windows.md

The absolute path is an example in documentation, not executable code, so no threat.

securityhigh SD-007 benign_example
scripts/deepstream/benchmark-ds.sh

The URL appears only in a comment block as a license reference, not executed by the script

permission_hygienehigh SD-003 benign_example
scripts/deepstream/ds-kitti-dump.sh

The content is documentation referencing path traversal concept, not actual traversal code.