The Agent Skills Directory

[SAFE]: The skill documents standard deployment procedures for the behavior-analytics service using Docker Compose. Instructions focus on environment setup, configuration mounting, and service verification.
[EXTERNAL_DOWNLOADS]: The skill facilitates downloading container images from the official NVIDIA Container Registry (nvcr.io). This is a well-known and expected service for this vendor.
[CREDENTIALS_UNSAFE]: Provides guidance for using NGC_CLI_API_KEY for registry authentication. Importantly, it includes a security note advising users NOT to hardcode the key, write it to shell profiles, or commit it to version control, recommending secure session-based input instead.
[COMMAND_EXECUTION]: Includes routine docker and docker compose commands for managing service containers. It also references a vendor-specific validation tool, nv-base, for skill evaluation.
[DATA_EXFILTRATION]: No patterns of unauthorized data collection or exfiltration were detected. Network operations are limited to interaction with well-known service brokers (Kafka, Redis, MQTT) and the official NVIDIA registry.
[PROMPT_INJECTION]: Instructions are purely operational and do not attempt to override agent safety guidelines or system behavior.

vss-setup-behavior-analytics