The Agent Skills Directory

[METADATA_POISONING]: The skill provides misleading information regarding Vitest versions and timelines. It claims Vitest 4.1.11 was released in August 2026 and describes specific security features (such as tighter mocker redirect handling) purportedly added in that version. As Vitest v4 does not exist at the time of analysis, this deceptive documentation could lead an agent to provide incorrect technical advice or misinterpret the security posture of a project.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze and review user-provided JavaScript and TypeScript code, which introduces a surface for instructions embedded in data to influence the agent.
Ingestion points: Ingests project source files, package.json, and vitest.config.ts files through the agent's file reading capabilities.
Boundary markers: Absent. The instructions do not include delimiters or specific commands for the agent to ignore potentially malicious instructions found within the project files it is reviewing.
Capability inventory: The skill has the ability to execute shell commands via pnpm for package installation and via the provided vitest-js-bootstrap.sh script.
Sanitization: No validation or sanitization of external content is performed before it is processed by the agent.
[DYNAMIC_EXECUTION]: The scripts/vitest-js-bootstrap.sh file utilizes a Python heredoc to generate and execute code at runtime. Although the current implementation is limited to formatting JSON, the pattern of assembling executable Python logic from environment variables at runtime is a dynamic execution vector.

vitest-js