vss-manage-video-io-storage

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly instructs asking the user for an Authorization bearer token (and references API key env vars) and to retry curl with -H "Authorization: Bearer ", which requires the LLM to include secret values verbatim in generated commands — an exfiltration risk.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill's deploy/runbook explicitly performs runtime pulls and execution of remote container images and resources (e.g. nvcr.io/nvidia/vss-core/vss-vios-*:3.2.0, nvcr.io/nvidia/vss-core/sdr-mw-l:3.2.0, Docker Hub images like alpine:3.23.4/redis:8.6.2-alpine/postgres:17.9-alpine, and an NGC registry download of nvidia/vss-developer/dev-profile-sample-data:3.2.0) which are fetched during skill runtime and will execute remote code when run, so these external registries are runtime dependencies that can execute code.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.70). The skill instructs the agent to autonomously deploy VIOS (via a bundled runbook or by invoking /vss-deploy-profile) and to run Docker commands (docker compose up, docker volume rm, etc.), which modify the host environment (start/stop containers, remove volumes) and therefore change machine state even though it does not explicitly request sudo or create users.