self-improve

Update skill documentation immediately based on learnings to prevent drift. Test commands before documenting them, document the why, and remove outdated info to keep OpenCode current and ensure all commands remain copy-paste ready.

Workflow automation self-improvement is an AI-learning capability that extends OpenCode by adding new skills, prompts, and guidance based on operational experience. It captures learnings from failed commands or discovered workflows to continuously evolve system capabilities.

Document the why behind new commands by recording the context and reasoning for each update directly in skill documentation. This maintenance rule ensures future AI-learning sessions understand the purpose behind workflow automation changes and prevents behavior drift.

Yes, you can fix outdated workflow automation commands by applying maintenance rules that enforce removing outdated info and testing commands before documenting them. This keeps OpenCode's skill documentation aligned with actual API behavior and reduces drift.

Trigger an OpenCode skill documentation update when API behavior differs, commands fail, or new workflows are discovered. Applying learnings immediately keeps the system current, enforces maintenance rules, and prevents documentation drift between behavior and reality.