The Agent Skills Directory

[Indirect Prompt Injection] (HIGH): The skill is designed to ingest and process untrusted external data from API responses, creating an attack surface for indirect injection.
Ingestion points: Data enters the system via Model.hydrate() and Model.collect() methods as shown in SKILL.md and references/models.md.
Boundary markers: There are no boundary markers or explicit instructions to ignore embedded commands implemented in the hydration flow.
Capability inventory: The booted() lifecycle hook in references/models.md demonstrates a dangerous pattern where registerPermissions(this.permissions) is called using data sourced directly from the input payload. This allows an attacker-controlled data source to potentially manipulate application security logic.
Sanitization: The provided examples lack any logic for validation, escaping, or sanitization of external content before it is processed by the model lifecycle.

nuxt-models