name: episode-start
description: Start a new learning episode in the self-learning memory system with proper context. Use this skill when beginning a new task that should be tracked for learning from execution patterns.

Episode Start

Start a new learning episode in the self-learning memory system.

Purpose

Create a new episode record with proper context for the memory backend to learn from execution patterns.

Steps

1. Understand the task: Parse the task description and identify:
2. Task type (implementation, debugging, refactoring, testing)
3. Domain (storage, patterns, retrieval, testing, etc.)

4. Language context (Rust/Tokio/async patterns)

5. Prepare TaskContext: Ensure you have:

6. language: "rust"
7. domain: One of [storage, patterns, retrieval, embedding, testing, ci]

8. tags: Array of relevant tags (e.g., ["turso", "async", "tokio"])

9. Create episode: Call SelfLearningMemory::start_episode(task_description, context)

10. Task description should be clear and concise (1-2 sentences)

11. Include relevant context from the user's request

12. Store episode_id: Keep the episode ID for logging subsequent steps

13. Initialize step logging: Prepare to log execution steps with:

14. Tool used
15. Action taken
16. Latency/tokens (if applicable)
17. Success status
18. Observations

Storage Requirements

• Persist to Turso (durable storage)
• Cache in redb (fast access)
• Store context as JSON blob

Example

let context = TaskContext {
    language: "rust".to_string(),
    domain: "storage".to_string(),
    tags: vec!["turso".to_string(), "async".to_string()],
};

let episode_id = memory
    .start_episode(
        "Implement async batch pattern updates",
        context
    )
    .await?;

Notes

• Always validate that both Turso and redb connections are healthy
• Use anyhow::Result for error handling
• Log any initialization failures with context