The Agent Skills Directory

[DATA_EXFILTRATION]: The skill performs network requests to https://api.ideogram.ai/generate to communicate with the Ideogram image generation service. This targets a well-known service consistent with the skill's primary purpose of providing an RBAC proxy for Ideogram.
[INDIRECT_PROMPT_INJECTION]: The skill processes external text input which is subsequently forwarded to a remote API.
Ingestion points: The prompt parameter in the teamGenerate function in SKILL.md is derived from user input.
Boundary markers: The prompt is interpolated into a JSON request body without specific delimiter-based boundary markers.
Capability inventory: The skill utilizes the fetch API to perform network POST operations.
Sanitization: The skill implements a checkContentPolicy function that uses regular expression blacklists to filter out restricted terms such as trademarked, violent, or explicit content before the prompt is processed.

ideogram-enterprise-rbac