The Agent Skills Directory

[COMMAND_EXECUTION]: The skill is instructed to automatically execute database migrations using the user's project-specific database driver and connection once the authentication code has been generated.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the user's project files during the initial stack detection phase. 1. Ingestion points: The agent scans local configuration files such as package.json, go.mod, and prisma/schema.prisma (Step 1). 2. Boundary markers: Not explicitly defined for the file scanning process. 3. Capability inventory: The skill has the capability to write multiple source code files (Step 5) and execute database migrations (Step 6). 4. Sanitization: While the skill includes rigorous validation rules for user inputs, it does not specifically sanitize the content of the project files used for stack detection.

create-auth