The Agent Skills Directory

[SAFE]: The skill documentation describes standard authentication workflows, including managed and headless modes, social logins, and role-based access control. All code examples use typical software development patterns.
[COMMAND_EXECUTION]: The skill includes examples for using a CLI tool (blink auth-config) to enable authentication providers. These are administrative commands intended for project setup and do not exhibit malicious intent.
[INDIRECT_PROMPT_INJECTION]: The authentication methods described (e.g., signUp, signInWithEmail) ingest untrusted user input such as email, password, and metadata. While this represents a data ingestion surface, the skill does not expose any dangerous capabilities (like shell execution or arbitrary file writes) that could be exploited via these inputs.
[DATA_EXPOSURE]: The skill correctly demonstrates using environment variables (VITE_BLINK_PUBLISHABLE_KEY) for project configuration. The hardcoded key in the example (blnk_pk_xxx) is clearly a placeholder and does not constitute a credential leak.

blink-auth