The Agent Skills Directory

[SAFE]: The skill documents standard integration patterns with Linear's API, a well-known service. It correctly identifies the use of pre-signed URLs for secure file uploads and provides examples of authenticated downloads using environment variables rather than hardcoded credentials.
[INDIRECT_PROMPT_INJECTION]: The skill describes functionality for processing external file uploads and URLs, which represents a potential attack surface. However, it includes proactive security guidance to mitigate risks from untrusted content:
Ingestion points: File paths (filePath) and external URLs (url) are ingested through tool arguments.
Boundary markers: Not present in the example code snippets, but the documentation provides architectural advice for validation.
Capability inventory: The skill is scoped to use Bash, Read, and Glob tools, alongside fetch for network operations to Linear's infrastructure.
Sanitization: The skill explicitly recommends content-type validation and EXIF metadata stripping to prevent malicious payloads.

Linear Attachments