The Agent Skills Directory

[SAFE]: The skill provides developer tools and integration guides for Sigma Auth. The scripts follow standard development practices for project initialization and configuration validation.
[EXTERNAL_DOWNLOADS]: The skill recommends installing @sigma-auth/better-auth-plugin and better-auth. These packages are legitimate dependencies for the integration, and @sigma-auth is the official namespace for the vendor 'b-open-io'.
[COMMAND_EXECUTION]: The skill utilizes bun to execute local scripts for project analysis (detect.ts), environment validation (validate-env.ts), and health checks (health-check.ts). These scripts perform necessary local file system reads and network connectivity tests to the authentication server.
[DATA_EXPOSURE]: The validate-env.ts script reads sensitive environment variables such as SIGMA_MEMBER_PRIVATE_KEY. The script is designed to validate the key's format (WIF) locally and explicitly avoids logging the sensitive values in its output report.
[INDIRECT_PROMPT_INJECTION]: The analysis tools (detect.ts) ingest data from local project files like package.json and source code. The scripts use basic string matching and JSON parsing to generate reports, which is a safe way to handle local untrusted project data without risk of influencing the agent's behavior.
[SAFE]: The automated scan flagged https://your-production-domain.com as a malicious URL. Analysis of the skill content confirms this string is used strictly as a placeholder in a documentation code block explaining how to configure CSRF protection (trusted origins) in Next.js, and it does not represent a functional URL or a phishing risk within the skill context.

setup-nextjs