The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands via pnpm to build project dependencies and run test suites using Vitest. These actions are standard for the skill's primary purpose of software verification.
[DATA_EXPOSURE_AND_EXFILTRATION]: The skill implements security best practices by using infisical for secret management. It includes explicit instructions to use the --silent flag and never print or echo secret values to the console.
[INDIRECT_PROMPT_INJECTION]: The skill interacts with external test specification files (.test.ts). This is an inherent attack surface for any test-running utility, but the instructions are focused on standard project testing workflows.
Ingestion points: Test files located in the specs/ directory (SKILL.md).
Boundary markers: None explicitly defined for file content.
Capability inventory: Command execution via pnpm and vitest (SKILL.md).
Sanitization: Relies on the standard execution environment of the Vitest runner.

run-tests