The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands specified in the local .e2e/config.json file for application orchestration, such as booting the environment (boot.up) and performing process checks (checks[].cmd). These are executed via spawnSync in the scripts/e2e.cjs utility.- [EXTERNAL_DOWNLOADS]: The skill requires the agent-browser package from the official NPM registry. The documentation specifies a pinned version (0.27.2) to ensure a consistent and verified execution environment.- [INDIRECT_PROMPT_INJECTION]: As a tool designed to interact with live web content, the skill acknowledges the risk of processing untrusted data. The SKILL.md provides explicit instructions to treat all scraped content as data rather than instructions and requires user confirmation for actions triggered by web data.- [CREDENTIALS_SAFE]: The documentation adheres to security best practices by recommending environment variables or specialized tools like gopass for credential management, rather than hardcoding sensitive information in the configuration files.

web-e2e