The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Shell tool to run test frameworks (Vitest, Jest) and search files using grep. It also executes dynamic logic via node -e to parse coverage summaries.
[EXTERNAL_DOWNLOADS]: The skill installs Node.js packages such as @vitest/coverage-v8 and other dependencies identified during test runs via pnpm. This introduces a potential supply chain risk if the agent is misled by malicious error messages into installing unauthorized packages.
[PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it processes code and configuration files from the repository. * Ingestion points: Reads package.json, config files, and all source/test files in the src directory. * Boundary markers: No delimiters or ignore instructions are used when reading file content. * Capability inventory: The skill can modify any file via the Write tool and execute shell commands via the Shell tool. * Sanitization: No sanitization is performed on ingested code or comments before they are used to guide refactoring or command execution.

test-coverage