The Agent Skills Directory

[SAFE]: The skill acts as a technical reference for BullMQ and does not contain any malicious instructions or safety bypasses.
[COMMAND_EXECUTION]: The skill mentions 'sandboxed-processors' in the Patterns section. This refers to a standard BullMQ architectural feature where jobs are executed in separate child processes for better isolation and reliability, and it is not used here for arbitrary command injection.
[PROMPT_INJECTION]: Regarding Indirect Prompt Injection (Category 8 surface analysis): The skill defines templates for workers that ingest job.data (SKILL.md). While the generic code snippets do not include explicit boundary markers or sanitization, this represents standard implementation patterns for the library rather than a malicious vulnerability in the skill itself. The worker capabilities involve database interaction via ioredis and typical network operations for tasks like sending emails.

bullmq-specialist