story-init
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill instructions are limited to gathering user input for a story project and creating standard markdown files and directories. No shell commands, network operations, or access to sensitive system paths were identified.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user data to populate story templates.
- Ingestion points: Story title, genre, synopsis, setting, and themes provided by the user in the initial workflow step.
- Boundary markers: Absent.
- Capability inventory: The agent creates directories and markdown files using the interpolated strings.
- Sanitization: Absent.
- Analysis: The risk is minimal as the ingested content is used solely for text-based documentation and is not processed as executable code, shell commands, or sensitive configuration.
Audit Metadata