The Agent Skills Directory

[SAFE]: The skill facilitates management of roles, teams, and collaborators on the TrueFoundry platform.
[SAFE]: Authentication is performed using standard environment variables (TFY_API_KEY and TFY_BASE_URL) or values from a .env file. The helper script tfy-api.sh implements a safe line-by-line parser for .env files to avoid executing arbitrary code or using shell expansion on untrusted input.
[SAFE]: Network operations are confined to the user-specified TFY_BASE_URL for interacting with the TrueFoundry API. The skill does not attempt to send data to unknown external domains.
[SAFE]: The helper script tfy-api.sh includes validation to prevent path traversal attacks by ensuring the API path starts with a forward slash and does not contain directory traversal sequences (e.g., '..').
[SAFE]: Privileged operations such as creating or deleting roles and adding collaborators are explicitly noted in the instructions as requiring human approval (HITL) and verification of identities.
[SAFE]: External references for container images target trusted and well-known registries including Amazon ECR (public.ecr.aws), GitHub Container Registry (ghcr.io), and NVIDIA's NGC catalog.

access-control