episode-start

Initialize a learning episode by creating a TaskContext with language 'rust', selecting a domain like storage or patterns, then calling SelfLearningMemory::start_episode() with your task description. The Skill prepares step logging and persists context to Turso while caching in redb for pattern tracking across executions.

Prepare a TaskContext specifying language as 'rust', choose a domain from storage, patterns, retrieval, embedding, testing, or ci, and add relevant tags. This structured context enables the memory system to learn from execution patterns in your specific domain.

Yes, the Skill is designed for Rust/Tokio/async context. It integrates with async task initialization, letting you start learning episodes within async workflows to track execution patterns across asynchronous operations.

episode-start initializes new episode records in a self-learning memory system that learns from execution patterns. By capturing domain, language, and tags upfront, it enables downstream analysis and retrieval of patterns across storage, testing, CI, and other domains.

After calling start_episode(), the context is persisted to Turso for durability and cached in redb for fast retrieval. An episode_id is stored and step logging is initialized, allowing you to track and learn from subsequent execution steps.