The Agent Skills Directory

Indirect Data Ingestion: The skill processes information gathered from external sources via a research subagent. This is an expected and functional part of the skill's purpose as a writing assistant. (1) Ingestion points: external findings are read from 'research/[slug].md' as specified in SKILL.md; (2) Boundary markers: no specific delimiting markers are used to separate research data from core instructions; (3) Capability inventory: the skill utilizes tool calls for image generation and task delegation, and performs file writing operations; (4) Sanitization: the skill does not explicitly perform sanitization on the ingested research findings. This data ingestion is well-contained within the intended research-to-writing pipeline.\n- Controlled File System Interaction: The skill organizes its work by writing to specific, task-appropriate directories ('blogs/' and 'research/'). This approach maintains a clear structure and ensures that file operations are confined to the relevant project scope.\n- Task Delegation: The use of a 'researcher' subagent for data gathering is a specialized architectural choice that allows the skill to focus on content synthesis and writing while leveraging platform-provided research capabilities.

blog-post