## What
Adds **[adelinamart/robrain](https://github.com/adelinamart/robrain)** to the 🧠 Knowledge & Memory section (alphabetical order, one server per line, Glama score badge included per the updated listing requirements).
## About the server
RoBrain is Apache-2.0 institutional memory for AI coding agents. Sensing (the MCP server) passively captures session turns; Perception extracts each decision into Postgres with a structured `rejected[]` field. At task time it surfaces the recorded rationale — including *why* an approach was rejected — before an agent steers down a path your team already ruled out. Runs entirely on your own machine (`npx robrain up`); nothing leaves your infrastructure.
- **Language/stack:** TypeScript · Postgres + pgvector · stdio transport (`npx -y robrain mcp`)
- **Works with:** Claude Code, Cursor, GitHub Copilot (VS Code), Codex CLI
- **Benchmarked:** VetoBench — 0/50 re-proposals of rejected approaches, receipts committed in `packages/vetobench/`
## Listing requirements
- ✅ Listed on Glama with **passing checks**: https://glama.ai/mcp/servers/adelinamart/robrain
- ✅ Score badge added after the repo link, using the required format
- ✅ Also published in the Official MCP Registry (`io.github.adelinamart/robrain`)
Supersedes my earlier PR (closed before the Glama listing existed).
