Screen prompts, skills and MCP tools with TypeSafe judgments and configurable reasoning models.
Set up and use Prompt Rejector to screen prompts, SKILL.md files, MCP tool descriptions, and agent capabilities for security risks. Use when the user requests Prompt Rejector installation, configuration, troubleshooting, or a security review of untrusted instructions. Includes cloning the repository, starting MCP or HTTPS, checking active TypeSafe judgments, and interpreting results.
