The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to guide an agent in running Vitest, which involves reading and executing code from test files. This creates a surface where malicious instructions could be embedded in the data being tested.
Ingestion points: The skill processes test files matching patterns such as **/*.test.ts, **/*.spec.ts, **/*.test-d.ts, and source files in src/**/*.ts (e.g., in references/features-filtering.md and references/advanced-type-testing.md).
Boundary markers: The documentation does not specify the use of delimiters or specific warnings to the agent to ignore embedded instructions within the test files it processes.
Capability inventory: The skill facilitates command-line execution (vitest CLI), file system writes (via snapshot testing and coverage reporting), and full code execution within the test environment (e.g., in references/core-cli.md and references/features-snapshots.md).
Sanitization: There are no mentioned mechanisms for sanitizing or escaping the content of the test files before they are processed by the agent or the test runner.

vitest