The Agent Skills Directory

[COMMAND_EXECUTION]: The skill involves running Maven commands (e.g., mvn test) to execute the generated Java tests. This is a primary and expected functionality for a testing-oriented skill and is limited to the local development environment.
[INDIRECT_PROMPT_INJECTION]: The skill processes project-specific data such as views and specifications to generate tests.
Ingestion points: Reads local Java source files and user-provided specifications.
Boundary markers: The instructions mandate a clear separation of concerns using the Actor/Page Object pattern, which acts as a structural delimiter.
Capability inventory: The agent can execute shell commands via Maven and write Java files.
Sanitization: The skill promotes translating business logic into structured Java code rather than direct interpolation, which reduces the surface for accidental instruction obedience.
[SAFE]: All external library references and URLs are directed towards official Vaadin resources, which are consistent with the skill's authorship and purpose. No malicious patterns such as obfuscation, persistence, or data exfiltration were detected.

vaadin-browserless-testing