The Agent Skills Directory

[SAFE]: The skill provides documentation and navigation for the NetBox platform. Analysis of the skill instructions and reference documents shows no evidence of malicious intent, data exfiltration, or unauthorized command execution.
[INDIRECT_PROMPT_INJECTION]: As a hub for network discovery and data ingestion tools such as Orb Agent and Diode, the ecosystem described has an attack surface for indirect prompt injection from network data. Ingestion points include external infrastructure data entering via Orb or Diode. Boundary markers and sanitization are addressed through referenced validation policies and compliance checks. Capabilities across the ecosystem include REST/GraphQL writes and automation via Ansible or Terraform.
[REMOTE_CODE_EXECUTION]: The documentation mentions support for custom scripts and Jinja2 templates within the NetBox application. It explicitly references security hardening, such as Jinja2 sandbox hardening in version 4.6.1, and the requirement for write-enabled tokens for execution, promoting secure usage of these features.

netbox