The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from user-provided OpenAPI specifications and discovery interviews to generate PHP source code (Controllers, Models, Policies) and database migrations. This creates a surface where a maliciously crafted OpenAPI file could attempt to influence the generated code output.
Ingestion points: openapi.yaml file and interactive design blocks in SKILL.md.
Boundary markers: No explicit prompt delimiters or 'ignore' instructions are used when interpolating spec data into code templates.
Capability inventory: The skill performs file-write operations for multiple PHP files and migration scripts.
Sanitization: The skill relies on template-based generation; explicit sanitization or validation of the input OpenAPI structure before code generation is not specified.
[COMMAND_EXECUTION]: The skill instructs users to perform standard development tasks by executing shell commands such as php artisan migrate and php artisan test. While these are routine for Laravel development, they involve the execution of local system commands based on the skill's guidance.

project-api