The Agent Skills Directory

[PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it analyzes untrusted data from writing samples, URLs, and file paths to create instructions that guide the agent's future output.
Ingestion points: Content is ingested through user-pasted text, WebSearch results, and local files accessed via the Read tool.
Boundary markers: The instructions do not specify the use of boundary markers or 'ignore' instructions when processing these external samples.
Capability inventory: The skill uses Read, Write, Edit, Grep, Glob, WebSearch, and AskUserQuestion tools.
Sanitization: There is no explicit sanitization or filtering logic mentioned for the provided writing samples.
[EXTERNAL_DOWNLOADS]: The skill references an external command or sub-skill /majestic-tools:interview. This is identified as a vendor-specific resource associated with the author 'majesticlabs-dev' and is used for its intended purpose of conducting a discovery interview.

brand-voice