The Agent Skills Directory

[SAFE]: The skill guides the agent through the legitimate process of setting up the CodeceptJS auth plugin for session management.
[SAFE]: The instructions explicitly emphasize security by forbidding hardcoded credentials and mandating the use of .env files with the Node.js process.loadEnvFile() API or dotenv package.
[SAFE]: It provides guidance on using the secret() wrapper for passwords, which is a security feature of the CodeceptJS framework to prevent secrets from appearing in logs.
[COMMAND_EXECUTION]: The skill mentions executing tests using npx codeceptjs run, which is the standard method for running CodeceptJS tests.
[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by instructing the agent to read HTML from a login page to identify locators.
Ingestion points: HTML content from the login page via run_code.
Boundary markers: No specific boundary markers or instruction-ignoring delimiters are mentioned for the ingested HTML.
Capability inventory: File system modifications to codecept.conf.js and steps_file.js, and shell command execution via npx.
Sanitization: No specific sanitization is performed on the ingested HTML, but the risk is low and inherent to the task of UI automation testing.

codeceptjs-auth