openagent-using-skills

Skill invocation governance works by intercepting user messages to check for applicable skills, loading relevant skill content, and directing behavior according to tool instructions before any reply is generated. This prevents unchecked responses.

You can prevent unchecked replies by mandating skill invocation across conversations. The system determines when a skill is relevant and loads it before responding, satisfying safety requirements and enforcing approval gates at phase transitions.

Pre-response skill checks are needed when your automated workflow requires strict safety and governance compliance. They enforce approval gates at phase transitions, stopping or requesting confirmation when needed to align actions with guidelines.

Yes, skill invocation enforcement provides platform-agnostic guidance by referencing the OpenCode/OpenAgent workflow. It applies across environments to determine skill relevance and load content before responding, ensuring consistent governance.

The best way to load relevant skills is to always invoke the Skill tool before any reply. This automatically checks for applicability, loads relevant skill content, and directs agent behavior according to tool instructions.

If an AI agent skips pre-response skill invocation, it risks responses that do not align with OpenAgent guidelines. The enforcement mechanism prevents this by mandating skill checks, highlighting applicable skills, and halting responses without proper loading.