The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses standard development commands including git for branch management, docker compose for environment orchestration, and Drupal-specific tools like composer and drush. These operations are essential for the skill's purpose of automating site builds and updates.
[EXTERNAL_DOWNLOADS]: The skill mentions checking for playwright-cli and potentially using npx to run @playwright/cli. This involves downloading and executing code from the official NPM registry, which is a well-known and reputable service for development tools.
[DATA_EXFILTRATION]: The skill includes instructions for posting reports to external platforms like GitHub, GitLab, or Slack. It adheres to safety best practices by explicitly requiring human approval before any such external transmission.
[PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes data from the website being tested and local project files.
Ingestion points: Site navigation menus, page titles, and project files like AGENTS.md.
Boundary markers: Data from the site is captured and placed into comparison reports without specific delimiters to separate it from instructions.
Capability inventory: The skill can execute shell commands (git, docker, drush) and write to the file system.
Sanitization: No specific sanitization of external site content is mentioned before it is processed or used in reports.

drupal-major-upgrade-validation