The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides standard command-line instructions for running end-to-end tests using the npx testcafe CLI.
[EXTERNAL_DOWNLOADS]: Dependencies are managed via standard package managers, fetching the official TestCafe framework from the npm registry.
[DATA_EXFILTRATION]: Instructions guide the user to set environment variables (LT_USERNAME, LT_ACCESS_KEY) for authenticating with the vendor's cloud testing platform at hub.lambdatest.com. This represents legitimate vendor functionality for remote test execution and does not constitute a security risk.
[INDIRECT_PROMPT_INJECTION]: As a code generation skill, it possesses an inherent attack surface for indirect injection.
Ingestion points: Processes user-provided requirements and website structures via the SKILL.md triggers.
Boundary markers: Utilizes markdown headers and fenced code blocks to delimit instructions.
Capability inventory: Includes shell command execution via npx testcafe and file system interaction for test creation.
Sanitization: Relies on the agent's internal logic for content generation without explicit instruction-level sanitization.

testcafe-skill