devflow:guide

DevFlow primitive discovery is performed by identifying all skills, sub-agents, hooks, and commands from the repository. The process scopes the plugin's skills directory to produce categorized groupings and a structured metadata summary for loading.

DevFlow discovery generates structured metadata fields including name, description, category, is_subagent, dependencies, components, and summary. These concise fields are structured specifically for vector embeddings to facilitate rapid loading.

Enumerating plugin commands and hooks involves running an automated discovery script across the repository. This produces structured groupings by category and a concise summary of totals, dependencies, and components to support developer onboarding.

Yes, the discovered DevFlow primitive metadata is structured for vector embeddings. It provides concise metadata fields including name, description, category, dependencies, and summary to ensure accurate indexing and retrieval.

Yes, DevFlow discovery identifies sub-agents separately from standard skills. The process includes an is_subagent metadata field in its structured output to distinguish agent status alongside categorized names and descriptions.

Grouping DevFlow skills by category is best handled through automated discovery scripts that scope the plugin's directory. This produces categorized groupings with a concise summary of dependencies and components for integration.