skill-authoring

To create reusable agent instructions, you define the skill's name, description, and operational steps within a structured markdown format. This approach saves workflows for future use, ensuring consistent task execution across independent sessions without needing to re-explain procedures.

Reusable agent skills are persistent instructions that extend an AI's capabilities for specific, multi-step workflows. You need them when you frequently execute identical complex tasks, like generating standardized reports, and want to trigger the process reliably without manual setup.

You write agent skills in markdown by adhering to specific frontmatter requirements for discoverability and operational integrity. This structured format defines the skill's name, description, and detailed instructions, enabling the agent to execute tasks accurately.

Yes, you can save multi-step workflows as reusable skills for persistent, session-independent execution. By defining your procedures in the skill's markdown instructions, the agent can reliably execute the entire workflow on command in future interactions.

Yes, specific frontmatter is required to make agent skills discoverable and operationally sound. Adhering to these structured markdown requirements ensures the system can properly identify and execute your saved instructions during an agent session.