creating-shot-list
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from story.md, characters.json, and backgrounds.json, creating an attack surface for indirect prompt injection.
- Ingestion points: Reads content from story.md, characters.json, and backgrounds.json (SKILL.md).
- Boundary markers: Lacks delimiters or warnings to ignore embedded instructions in source files.
- Capability inventory: Operations are restricted to local file reading and writing shots.json. No network or shell execution tools are used.
- Sanitization: Input data is interpolated into JSON fields without filtering or escaping.
- [SAFE]: No malicious patterns, obfuscation, or dangerous system commands were detected. The skill mandates human-in-the-loop review.
Audit Metadata