The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to generate prompts that interpolate {{user}} data into the agent's context. This is a standard attack surface for storytelling agents where malicious input could attempt to influence the narrative logic.
Ingestion points: {{user}} variables are integrated into the storyline_system_prompt.txt, character_cards.txt, and RPG mechanics descriptions in SKILL.md.
Boundary markers: Absent. The templates use logical headers (e.g., [WORLD & SCENARIO RULES]) but do not include specific delimiters or instructions to ignore commands embedded in user-supplied text.
Capability inventory: The skill itself does not request any sensitive tool access or perform dangerous system operations in SKILL.md.
Sanitization: Absent. There is no explicit logic provided in the templates to filter or sanitize user input before it is processed by the agent engine.
[SAFE]: The skill references legitimate external domains associated with the ISEKAI ZERO platform (isekaizero.ai, docs.isekaizero.ai). No obfuscation, hardcoded credentials, or unauthorized network operations were found. The mentioned Python script (generate_isekaizero_package.py) is a local helper utility described for content packaging and does not involve remote code execution from untrusted sources.

isekaizero-storyline-architect