The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a project validation utility in scripts/check-versions.sh which uses shell commands to verify the environment.
The script utilizes node -p to evaluate short JavaScript snippets for version extraction from package.json files.
It employs the find utility to scan the project directory for deprecated file patterns and verify parallel route structures.
[EXTERNAL_DOWNLOADS]: The documentation references official migration tools that involve remote execution.
The migration guide recommends running npx @next/codemod@canary upgrade latest, which fetches and executes code from the official Next.js package repository.
This tool is provided by a well-known service and a trusted organization for project maintenance.
[INDIRECT_PROMPT_INJECTION]: The skill operates on user-provided source code, creating an attack surface for indirect instructions.
Ingestion points: The skill reads project source files and configuration data via Read and Grep tools.
Boundary markers: None are explicitly used in the templates to separate instruction from processed code data.
Capability inventory: The skill possesses file modification (Write, Edit) and shell execution (Bash) capabilities.
Sanitization: While the templates demonstrate secure practices like URI encoding in server actions, no explicit sanitization is applied to the input data the agent processes.
[SAFE]: The skill follows security best practices, including the use of httpOnly and secure flags for cookies in its API templates. No indicators of data exfiltration, credential theft, or persistence were found.

nextjs