ai

Centralize project agent skills by hosting all SKILL definitions under the .ai/skills directory, which ensures IDEs and tools reference a single source of truth for consistent execution. This standardizes skill definitions across your project environment.

Scaffold code by asking the AI to apply the appropriate skill from the .ai/skills store. The AI automatically selects and executes the matching skill to generate or scaffold code modules, enforcing a single source of truth for skill definitions.

Automated skill selection works through read-and-apply semantics: when you request to add a module, the AI reads the canonical SKILL.md file from the centralized .ai/skills store and runs the corresponding workflow to generate the code.

Yes, you need a .ai/skills directory structure to host the canonical SKILL.md files. This directory acts as the centralized store that IDEs and tools reference to maintain a single source of truth for project agent skills.

Yes, you can run predefined workflows that rely on skills defined under .ai/skills. The AI applies the appropriate skill from the canonical store to execute the workflow, ensuring all tools reference the same standardized skill definitions.

IDEs reference inconsistent skill definitions when they are not centralized. By hosting all SKILL.md definitions under a single .ai/skills directory, you enforce a single source of truth, ensuring IDEs and tools reference the canonical skill definitions consistently.