Recall Reasoning

This searches your past Claude sessions to surface what worked and what didn't. It queries two indexes: artifact handoffs with post-mortems (successful patterns, dead ends, key decisions) and reasoning files from build attempts. The workflow is straightforward: search the artifact index first for rich context, then drill into reasoning files if you need specifics on build errors. The real value is in those post-mortem sections that explicitly mark approaches as succeeded or failed, so you're not rediscovering the same dead ends. If you're working on something similar to a past session or debugging a familiar issue, this beats scrolling through commit history.