The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data (existing Cypress or Selenium test files) and uses the content to generate new executable code. This creates a surface where malicious instructions embedded in the source tests could influence the agent's behavior.
Ingestion points: The skill reads Cypress/Selenium test files and configuration files (e.g., package.json, cypress.config.ts) as described in SKILL.md (Steps 1 and 4).
Boundary markers: Absent; there are no specific instructions to the agent to ignore or delimit embedded natural language instructions within the test code.
Capability inventory: The skill performs file writes (converting files) and executes shell commands via npx playwright test (Step 7) and potentially modifies CI workflows (Step 8).
Sanitization: Absent; the skill directly maps and translates code blocks without explicit validation or escaping of the input content.
[COMMAND_EXECUTION]: The skill uses the shell to execute testing commands to verify the migration success.
Evidence: SKILL.md (Step 7) explicitly instructs the agent to run npx playwright test <converted-file> --reporter=list.
[DYNAMIC_EXECUTION]: The skill generates new test scripts (JavaScript/TypeScript) at runtime and immediately executes them to verify functionality.
Evidence: The conversion process in SKILL.md (Step 4) followed by the verification in Step 7 represents a script generation and execution workflow.

migrate