The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses scripts/vss-bootstrap.sh to retrieve the application source code from the author's official repository at https://github.com/open-edge-platform/edge-ai-libraries.git. The script uses security-conscious git flags, including shallow and sparse checkouts, to minimize the amount of data fetched.
[COMMAND_EXECUTION]: The skill automates the application lifecycle via shell scripts and Docker Compose commands. It includes robust operational checks, such as verifying the Docker daemon's status before attempting deployment and monitoring health endpoints to confirm successful startup.
[INDIRECT_PROMPT_INJECTION]: The skill handles user-provided configuration, such as deployment modes and port numbers. It manages the risk of accidental or malicious parameter manipulation by explicitly instructing the agent to ground all commands in the repository's valid files and by employing proper shell-quoting mechanisms (%q) in the gen-secrets.sh utility to sanitize environment variables.
[SAFE]: The skill's credential management system generates secure, randomized passwords for internal services like PostgreSQL and MinIO. These are stored locally in a configuration file with restricted filesystem permissions (mode 0600), ensuring that sensitive data is neither hardcoded nor exposed in the workspace.

vss-deploy