soap-opera-sequence
Pass
Audited by Gen Agent Trust Hub on Jun 29, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No attempts to override system instructions or bypass safety filters were detected. The instructions are strictly focused on generating marketing copy.
- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill does not access sensitive local files or environment variables. It contains one external link for attribution to 'copyaccelerator.com', which is the official site for the RMBC framework described in the skill.
- [EXTERNAL_DOWNLOADS]: No remote scripts or binary downloads are present. The skill references a local file 'rmbc-context/SKILL.md' which is part of the skill's internal directory structure.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input via fields like 'product_description' and 'origin_story_hook'. While it lacks explicit delimiters or boundary markers to prevent the AI from following instructions embedded in those inputs, the skill's lack of dangerous capabilities (no shell access, no network exfiltration, no file-system writing) limits the potential impact to content corruption only.
- Ingestion points: Inputs defined in the frontmatter (SKILL.md).
- Boundary markers: Absent.
- Capability inventory: Limited to text generation; no system or network tools allowed.
- Sanitization: No specific sanitization or filtering is applied to the inputs.
- [OBFUSCATION]: The content is clear and uses standard Markdown formatting with no hidden characters, encoding, or deceptive text patterns.
Audit Metadata