The Agent Skills Directory

[Indirect Prompt Injection] (HIGH): The skill is susceptible to indirect prompt injection because it processes untrusted external data (project files) and has capabilities to modify the filesystem.
Ingestion points: The skill uses Read, Glob, and Grep tools to pull external file content into the agent's context.
Boundary markers: No delimiters or explicit instructions to ignore embedded commands within the analyzed code are provided.
Capability inventory: The skill utilizes Write and Edit tools, granting it the power to alter the workspace based on its analysis.
Sanitization: There is no logic present to sanitize or escape instructions found within the code being reviewed.
[Data Exposure & Exfiltration] (SAFE): The skill correctly identifies environment variables and secrets management as critical security principles but does not attempt to access or exfiltrate sensitive files itself.
[Unverifiable Dependencies & Remote Code Execution] (SAFE): While the skill mentions various npm packages (e.g., Fastify, Zod, Helmet), it does so for architectural advice and does not perform runtime installations or remote script executions.

nodejs-best-practices