Your coding agent remembers more than MEMORY.md — so I built a local inspector for it

Developer and creator 'hailneed' has introduced AgentMemora, an open-source tool designed to help developers manage and inspect the fragmented context generated by AI coding agents. As AI agents often scatter information across various sessions, subagent transcripts, and local files, it becomes difficult for developers to maintain continuity when starting new sessions. AgentMemora acts as a local inspection layer that organizes these disparate data points by project, allowing users to curate, preserve, and transfer essential context. The tool currently supports Claude Code and provides a dashboard to identify stale memory, export 'Context Capsules,' and perform 'Context Surgery' to extract only relevant decisions and constraints. By keeping data local and avoiding external APIs or telemetry, the project aims to solve the problem of context loss in long-running AI development workflows. The tool is available via npm and is currently seeking community feedback to refine its features.