The Agent Skills Directory

[SAFE]: The skill serves as an integration guide for the shadcn/ui component library. It includes standard documentation and React component examples.
[COMMAND_EXECUTION]: The skill includes a local utility script scripts/verify-setup.sh which performs benign environment checks using standard tools like grep, find, and wc to validate project configuration files such as components.json, tailwind.config.js, and tsconfig.json.
[INDIRECT_PROMPT_INJECTION]: As a development assistant, the skill has an ingestion surface for untrusted data by reading project source files and configuration.
Ingestion points: Reads project configuration files (components.json, package.json, tsconfig.json) and source files (globals.css, .tsx files) via the Read tool and validation script.
Boundary markers: None explicitly defined in the provided instruction set.
Capability inventory: The skill has Write and Bash capabilities to modify code and run setup verification.
Sanitization: No specific sanitization or escaping of file content is defined in the instructions.
Note: This surface is inherent to the skill's primary purpose of code integration and is not considered a finding that escalates the verdict.

shadcn-ui