worldbuilding
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an inherent attack surface for indirect prompt injection because it processes untrusted story data and user input to generate project files. \n
- Ingestion points: The skill reads
story.md,worldbuilding/_index.md, and conversational user input to define world elements (SKILL.md). \n - Boundary markers: Absent; the instructions do not provide delimiters or specific guidelines to help the agent distinguish between data and potentially malicious instructions within the story files. \n
- Capability inventory: The skill utilizes file system read and write capabilities, specifically for creating and updating markdown files within the
worldbuilding/directory and related character files (SKILL.md). \n - Sanitization: Absent; there is no mention of content filtering, escaping, or validation of the text retrieved from story documents. \n- [SAFE]: The skill's behavior is entirely consistent with its stated purpose of assisting in story development. All file operations are localized to the user's project structure, and no suspicious external dependencies, obfuscation, or persistence mechanisms were found.
Audit Metadata