The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided business briefs and external websites as design references, creating an attack surface for indirect prompt injection.
Ingestion points: Plain-language business briefs and content from external reference websites (e.g., must-have pages, assets, or reference sites).
Boundary markers: Absent. The skill does not specify the use of delimiters to separate untrusted user input from its own instructions.
Capability inventory: The skill can create companies (projects.create), execute arbitrary API operations (execute_operation), manage tasks (tasks.create), and publish live websites (apps.publish).
Sanitization: No sanitization or validation steps are described for the input data before it is used in task payloads.
[EXTERNAL_DOWNLOADS]: The skill connects to an external API at api.doany.ai/mcp and directs users to external documentation and CLI guides hosted on docs.doany.ai.
[COMMAND_EXECUTION]: The instructions describe using a terminal and a command-line interface to perform equivalent API operations if the primary MCP endpoint is unavailable.

website-builder