reflexion

To capture task feedback for pattern discovery, you log session details such as the task description, reward, success rating, and critique. This records episodes of work to train the recommendation system and refine future suggestions.

Storing feedback for experiential learning requires a session_id, task, reward, success, and critique parameters. These inputs are logged using the store operation to persist experiences for the recommendation system.

Logging post-work critique helps refine future recommendations by capturing what worked and what didn't. This feedback data enables the learner to auto-discover new patterns over time and improve suggestion accuracy.

Yes, you can retrieve past session feedback to review workflow patterns. The skill exposes retrieve operations to manage and access previously stored feedback data, supporting continuous improvement during workflow reviews.

The best way to record episodes of work for a recommendation system is to use the reflexion_store function. This logs the task description and success rating after completing a task to support continuous pattern improvement.

You should use feedback capture to support continuous improvement after completing a task or during workflow reviews. Logging what worked and what didn't provides the data necessary for the learner to auto-discover new patterns.