glide-mq
Warn
Audited by Gen Agent Trust Hub on Mar 19, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of the 'glide-mq' package from the npm registry. This package is an external dependency whose contents are not verified within the skill itself.
- [COMMAND_EXECUTION]: Instructions in the documentation lead users to execute 'npm install glide-mq', which results in the installation and potential execution of third-party code on the host system.
- [CREDENTIALS_UNSAFE]: The connection example in 'SKILL.md' contains a hardcoded password value ('secret'). While likely intended as a placeholder, it follows a pattern that can lead to credential exposure if developers use it literally.
- [PROMPT_INJECTION]: The skill provides patterns for message workers that ingest data from external job queues, creating an attack surface for indirect prompt injection. 1. Ingestion points: The 'Worker' implementation in 'SKILL.md' receives and processes the 'job.data' object directly. 2. Boundary markers: No delimiters or 'ignore-previous-instructions' warnings are used when processing the data. 3. Capability inventory: The worker callback can perform any system or network operation, though the example only shows logging. 4. Sanitization: No input validation, escaping, or sanitization of the job data is present in the provided examples.
Audit Metadata