The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user data via multiple input fields and interpolates them directly into the image generation prompt without using sanitization, escaping, or boundary markers. This creates a surface where adversarial user input could attempt to influence the image generation model's behavior.
Ingestion points: SKILL.md (inputs: brand_name, style, industry, colors, background).
Boundary markers: None identified in the prompt template to delimit user content.
Capability inventory: The skill has the capability to execute the muapi CLI and perform network operations via curl to the vendor's API.
Sanitization: No input validation or filtering is implemented for the user-supplied strings.
[COMMAND_EXECUTION]: The skill's operational instructions direct the agent to execute shell commands using the muapi CLI and provides a fallback curl command for interacting with the MuAPI web service.
Evidence: Use of muapi image generate, muapi auth configure, and curl -X POST https://api.muapi.ai/api/v1/... in the agent instructions.

muapi-logo-generator