The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting untrusted user inputs and interpolating them into prompts for downstream generation tools.
Ingestion points: User-provided inputs including product_name, campaign_message, target_audience, and visual_style are defined in SKILL.md.
Boundary markers: The skill does not employ explicit boundary markers or delimiters when inserting these inputs into tool prompts.
Capability inventory: The skill uses the muapi CLI and curl to perform network operations and generate content via external APIs.
Sanitization: There is no evidence of sanitization or validation of the input strings before they are used in commands or prompts.
[DATA_EXFILTRATION]: The skill performs network operations using curl to the api.muapi.ai domain. These operations are used to interface with the vendor's API for content generation and involve the use of the MUAPI_API_KEY for authentication, which is consistent with the skill's primary functionality.

muapi-product-campaign