Cc-reaper – Three-layer cleanup for orphan Claude Code processes
Three-layer orphan process cleanup for Claude Code's known memory leak.

MCP server pipes context to Claude Code so it stops hallucinating architecture.
Engineering teams using Claude Code or Cursor
Sourcegraph Cody · Cursor · Notion
You can have a memory layer where you store everything about your company or project. When an AI agent is used by an engineer, you can simply ask it to use the CLI, and it will go and retrieve the relevant information.
There is also management of who last edited a file, so you can understand who is responsible for the latest context level changes stored in Brifly.
You can create different spaces, invite people to them, create threads for discussions, and add video and audio messages. By the way, transcriptions from video messages are also taken into account by the AI chat when you ask questions.
I believe this is a great tool for teams, or even for solo users, especially when you have a lot of information and need to keep it updated all the time.
There are also other cool features that can be useful. It is completely free, and if you need AI features, just add your OpenAI api key.
Three-layer orphan process cleanup for Claude Code's known memory leak.
Dependency graph persists across AI sessions; Claude never rescans the same files twice.
Claude Code stop hook that keeps agents working when you step away from your desk.
Finally gives Claude Code persistent log memory without manual copy-pasting.
Bidirectional code↔diagram with Claude MCP beats text-only codebase analysis tools.
SQLite-vec + tree-sitter chunking cuts Claude Code costs by 39% versus grep-based search.