The Agent Skills Directory

Indirect Prompt Injection (LOW): The skill is designed to ingest and analyze external source code and documentation, which is an inherent attack surface for indirect prompt injection if the ingested data contains malicious instructions.
Ingestion points: Source code files and external documentation referenced during the research phase.
Boundary markers: No specific boundary markers or 'ignore' instructions for external data are defined.
Capability inventory: Capability to write documentation files to local directories like docs/ or ai_docs/.
Sanitization: No explicit sanitization or validation of the ingested content is mentioned.
NO_CODE (SAFE): The skill does not contain any executable scripts, binary files, or configuration files that could execute code. It is purely composed of instructional markdown.

tech-blog