The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted script data to extract structured information, creating a potential vector for indirect prompt injection if the source scripts contain malicious instructions.
Ingestion points: External script content is ingested via the read_script_for_extraction tool as defined in SKILL.md.
Boundary markers: The instructions lack explicit boundary markers or delimiters to differentiate between system instructions and the script content being analyzed.
Capability inventory: The skill has the ability to write to character, scene, and prop databases using the save_dedup_characters, save_dedup_scenes, and save_dedup_props tools.
Sanitization: No explicit sanitization, validation, or escaping logic is described for the content extracted from the scripts before it is stored in the project database.

extractor