The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process asset manifest data provided by third-party generators. This creates a surface for indirect prompt injection if the external data contains malicious instructions meant to influence the agent's behavior during the manifest generation or handoff phases.
Ingestion points: Asset manifest rows provided by third-party generators (as described in the 'Output and ownership' section).
Boundary markers: The instructions do not specify any delimiters or safety markers to isolate untrusted manifest data from the agent's control instructions.
Capability inventory: The skill generates asset manifests (file writing) and coordinates with other skills such as sprite-atlas and asset-pack-intake (skill chaining).
Sanitization: There is no evidence of validation or sanitization of the input data from third-party sources.

sprite-catalog