The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands to run tests via npx playwright test and capture output using tee and ls. These are standard developer operations for CI/CD and local testing workflows.- [DATA_EXFILTRATION]: The skill is designed to send test summaries to external endpoints using curl. This behavior is controlled by user-defined environment variables such as SLACK_WEBHOOK_URL, making the communication intentional and configurable by the user.- [SAFE]: The skill processes external data from test logs and JSON results, which presents an inherent surface for indirect prompt injection. However, this is expected behavior for a reporting tool. Evidence chain: (1) Ingestion points: test-results/ and test-output.log. (2) Boundary markers: Absent; test data is interpolated directly into templates. (3) Capability inventory: Network access via curl, command execution via npx, and file writes to test-reports/. (4) Sanitization: None identified for test names or error messages.

report