world-bible
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [NO_CODE]: The skill consists exclusively of markdown instructions for manual world-building workflows. It does not include any scripts, executables, or third-party package dependencies.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow that ingests external data from story briefs and architecture (SKILL.md). It lacks formal boundary markers or sanitization for these inputs. However, the identified capabilities are limited to writing text-based rules and location descriptions to
01_story/world.md. This ingestion surface is considered safe as it involves no code execution or network communication. - [SAFE]: No malicious patterns, obfuscation, or persistence mechanisms were detected in the skill metadata or workflow instructions.
Audit Metadata