The Agent Skills Directory

[PROMPT_INJECTION]: The skill contains instructions designed to programmatically bypass user interaction requirements in secondary tools. In Phase 2 of SKILL.md, it explicitly directs the agent to drive the story-long-analyze pipeline in a mode that hits a 'skip asking' path, effectively suppressing manual checkpoints and confirmations that would otherwise be presented to the user.
[PROMPT_INJECTION]: The skill's primary function involves processing untrusted novel text provided by users, which introduces a surface for indirect prompt injection. Malicious instructions hidden within imported books could attempt to influence the agent's behavior during analysis or project generation.
Ingestion points: Untrusted data enters the agent context through file paths or raw text pasting as defined in Phase 1 of SKILL.md.
Boundary markers: There are no explicit instructions or delimiters used to wrap the imported text to warn the agent against executing embedded instructions.
Capability inventory: The skill possesses significant capabilities, including extensive file system access (creating/modifying directories and files) and the ability to trigger and influence other agent skills and sub-agents such as chapter-extractor and story-explorer.
Sanitization: The instructions do not define any sanitization, escaping, or validation steps for the content being imported from user-provided sources.

story-import