The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions include standard development commands for executing tests and checking coverage locally.\n
Evidence: npm test and npm run test:coverage in SKILL.md.\n- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for reading and analyzing external source code to generate unit and integration tests. This processes potentially untrusted data that could contain malicious instructions.\n
Ingestion points: Frontend source code files in the target directories (SKILL.md, references/workflow.md).\n
Boundary markers: Relies on the agent performing a complete reading of the source file to understand context, rather than using explicit delimiters.\n
Capability inventory: Shell command execution via npm test is documented for verifying generated code.\n
Sanitization: Mitigation is provided through the use of standardized templates and a verification checklist to ensure test quality and correctness.

frontend-testing