The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill explicitly asks the user for their email and password in Step 2 if authentication is required to access the app's features. While this is used to automate the login process via Playwright, it involves the handling of sensitive clear-text credentials by the agent.
[COMMAND_EXECUTION]: The skill executes shell commands to check for Playwright's installation, create directories, and run the generated Node.js script. It also uses sips or file to verify the generated images.
[DYNAMIC_EXECUTION]: In Step 6, the skill generates a temporary Node.js script (screenshot-script.mjs) containing user-provided data (URL, credentials, and paths) and then executes it using node. This dynamic script generation and execution pattern is a common vector for command injection if inputs are not properly sanitized.
[INDIRECT_PROMPT_INJECTION]: The skill reads external files such as README.md, CHANGELOG.md, and routing configurations to identify features for screenshots. This creates a surface for indirect prompt injection if those files contain malicious instructions designed to influence the agent's behavior during the analysis phase.
Ingestion points: Reads README.md, CHANGELOG.md, docs/, and framework-specific routing files (e.g., config/routes.rb, app/ directory, src/router/index.js).
Boundary markers: None identified; the skill instructions do not specify using delimiters or 'ignore' instructions for the content of these files.
Capability inventory: Uses mkdir, node (to execute generated script), rm, ls, and image identification tools (sips/file).
Sanitization: The skill does not explicitly mention sanitizing or escaping the content read from the documentation or route files before using it to plan screenshots or build the feature list.

screenshots