The Agent Skills Directory

[UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: The skill installs the explorbot package via npm and utilizes npx playwright install to download browser binaries. These are standard functional requirements for the testing framework and originate from the vendor or well-known service providers.\n- [COMMAND_EXECUTION]: Multiple shell commands including npm, npx, curl, and grep are used to manage project settings, install dependencies, and verify connectivity to the target application. These commands are integral to the setup process.\n- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill interacts with sensitive file paths including the project-local .env file and a global configuration file at ~/.explorbot/.env. It mitigates risk by instructing the agent to verify API key presence via grep rather than echoing values, and confirms that .env files are excluded from version control via gitignore.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided application URLs and navigation paths, which are then interpolated into verification commands (curl, npx explorbot navigate). While this creates a potential surface for injection, the instructions provide guidance on verifying these inputs from project files first.

explorbot-setup