The Agent Skills Directory

[SAFE]: The skill operates as a high-level creative assistant for content transformation. It utilizes standard instruction-following capabilities to rewrite text into specific templates and uses declared Notion integrations to store the results. No evidence of obfuscation, credential harvesting, or unauthorized network access was found.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external, untrusted "source material" to generate scripts, which creates a potential surface for indirect prompt injection.
- Ingestion points: Untrusted data enters the agent context through the "source material" provided by the user for transformation (Step 1 of the algorithm).
- Boundary markers: The instructions lack specific boundary markers (e.g., XML tags) to wrap the untrusted input, though it includes a checklist to ensure the output matches the viral style.
- Capability inventory: The skill has the capability to write to external services (creating Notion pages) based on the processed content.
- Sanitization: There is no explicit sanitization of the input data; however, the strict requirement to rewrite the content into "viral formats" based on specific reference examples acts as a functional transformation layer that reduces the risk of the agent executing instructions embedded within the source text.

reels-creator