The Agent Skills Directory

[SAFE]: The skill provides high-quality documentation, templates, and utility scripts for Clerk authentication. It proactively addresses security concerns, such as the March 2025 Next.js middleware CVE, by instructing users on proper configuration changes.
[DATA_EXPOSURE_AND_EXFILTRATION]: Analyzed for credential theft. The skill uses standard development placeholders (sk_test_..., pk_test_...) in examples and templates. It provides clear guidance on using environment variables and secret management tools (like Cloudflare Secrets) to prevent sensitive data exposure. Network operations are strictly limited to official Clerk and Svix APIs for authentication and webhook verification.
[EXTERNAL_DOWNLOADS]: Analyzed for risky dependencies. The skill references well-known and trusted packages from the official NPM registry (e.g., @clerk/nextjs, svix, hono). All external documentation links point to reputable sources including Clerk, GitHub, and StackOverflow.
[INDIRECT_PROMPT_INJECTION]: Analyzed for exploitable data surfaces.
Ingestion points: The clerk-setup agent uses grep and find to read local project configuration files like package.json and .env to assist with setup.
Boundary markers: The agent uses explicit workflow headers and a 'Do NOT' section to define clear operational boundaries.
Capability inventory: Uses Bash, Read, and Edit tools to perform setup tasks and troubleshooting.
Sanitization: The skill generates informational reports and configuration templates rather than executing ingested data as code.

clerk-auth