The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses kubectl to modify environment variables and manage deployment rollouts in Kubernetes clusters. This involves high-privilege cluster access but is directly related to the skill's primary function of system migration.
[EXTERNAL_DOWNLOADS]: The skill initiates the installation of the @ideogram/sdk package via npm. This is a standard dependency for a well-known service and is documented neutrally as a functional requirement.
[PROMPT_INJECTION]: The skill performs analysis on the local codebase using find and grep, creating a potential surface for indirect prompt injection if those local files contain adversarial content designed to mislead the AI agent. * Ingestion points: Local source code files (.ts, .py) and package listings. * Boundary markers: Absent; the skill processes local file content without specific delimiters or instructions to ignore embedded directives. * Capability inventory: Extensive system access including Bash (npm, node, kubectl) and Write/Edit permissions. * Sanitization: None; the content is retrieved and processed as raw text.

ideogram-migration-deep-dive