Self Reflection

This lets your AI agent track its own mistakes and lessons learned over time through a persistent reflection log. You can trigger it manually or via heartbeat to record what went wrong, what was learned, and how to improve next time. It's a straightforward approach to giving agents something like institutional memory across sessions. The commands are basic (check, log, read, stats) and it's already got 129 installs, which suggests people find value in making their agents slightly less prone to repeating the same errors. Worth trying if you're running longer workflows where the same patterns keep tripping things up.