The Agent Skills Directory

[SAFE]: The skill is an architectural and security audit tool designed to help developers maintain correct React Server Component boundaries. It does not perform any suspicious network operations or access sensitive data.
[COMMAND_EXECUTION]: The skill provides numerous shell one-liners and a bash script (validate-rsc.sh) that use standard utilities such as grep, ripgrep, jq, head, sed, and awk. These commands are strictly used for read-only inspection of the project's source code and configuration files. There are no patterns suggesting arbitrary command injection or privilege escalation.
[PROMPT_INJECTION]: The instructions are focused on guiding the agent through a specific validation workflow. While it uses imperative language like 'ALWAYS' and 'NEVER', these are technical guidelines for the auditing process and do not attempt to bypass AI safety filters or extract system prompts.
[DATA_EXPOSURE_AND_EXFILTRATION]: No network exfiltration patterns were found. The skill does not access sensitive file paths (e.g., .env, .ssh, .aws) or attempt to transmit data to external domains. Documentation links provided are all to official project websites (react.dev, nextjs.org, shadcn.com).
[EXTERNAL_DOWNLOADS]: The skill contains a reference to a GitHub Action workflow in references/methods.md that mentions sudo apt-get install -y ripgrep jq. This is standard documentation for configuring a CI/CD environment and does not constitute a runtime risk for the user.

shadcn-agents-rsc-boundary-validator