The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external novel content, creating a potential surface for indirect prompt injection attacks.
Ingestion points: Novel content is ingested via the read_episode_script tool mentioned in SKILL.md.
Boundary markers: There are no instructions to use delimiters or ignore instructions found within the input text.
Capability inventory: The skill references the save_script tool for file writing but does not have network access or shell execution tools.
Sanitization: The provided instructions do not include steps to sanitize or validate the input data.
[NO_CODE]: The skill consists of markdown instructions only and does not contain any scripts, binaries, or executable code.

script-rewriter