What problem does it solve? Hard-won debugging insights and workflow lessons often disappear after a task ends, forcing future agents to rediscover the same traps. This Skill turns verified experience into retrieval-friendly guidance that shapes future Just Demand behavior. ## Core Features & Use Cases - Destination Selection: Decide whether a lesson becomes a new skill, extends an existing skill, or stays archived with the task. - Structured Capture Format: Write lessons with trigger conditions, failure modes, and preferred actions so future agents can retrieve and apply them quickly. - Boundary Enforcement: Prevent raw logs, secrets, unverified guesses, and one-off details from polluting reusable skills. - Use Case: After a long debugging session reveals a recurring state-machine trap, use this Skill to encode the trigger, symptom, and fix as a new skill under .opencode/skills/ so the next agent avoids the same mistake. ## Quick Start After verification passes on a non-trivial debugging task, ask the agent to capture the lesson learned as a reusable skill or archive it with the task.
