The Agent Skills Directory

[COMMAND_EXECUTION]: The skill requires the file_write capability to generate the DESIGN.md and brief-preview.html files. It also generates functional HTML and CSS code based on user input for the platform's preview feature, which involves the dynamic creation and execution of code within the agent's environment.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it processes untrusted user data from the brief input and incorporates it into generated files and visual previews without explicit sanitization.
Ingestion points: The brief variable in SKILL.md which accepts natural language or I-Lang protocols.
Boundary markers: Absent; there are no instructions to use delimiters or ignore instructions embedded within the user input.
Capability inventory: The skill utilizes the file_write capability across its workflow.
Sanitization: No sanitization, escaping, or validation of the brief input is defined before it is used to generate the final documentation and preview code.

design-brief