Shared memory for researchers

Research accumulates context that is painful to reconstruct: the papers you already read and why they did or did not matter, the dead ends you ruled out, the methods that worked, the open questions, the precise definitions a project depends on. When you use AI agents to search literature, summarize papers, or reason about findings, they start each session without that history, so you re-feed the same background and they suggest leads you exhausted weeks ago. Glen, shared memory for AI agents, gives the agents researchers use long-term shared memory as a single MCP tool, so your accumulated knowledge persists and is shared across collaborators.

Connect your research tools to Glen over MCP and any agent, surveying literature, summarizing a paper, or reasoning about a result, can call one tool that retrieves relevant long-term context and records new facts in a single round trip. Before it works, it pulls what the project already knows: the sources already assessed, the hypotheses tested, the methods chosen, the conclusions reached. As you read and reason, it writes back new findings and judgments, so the next session, yours or a collaborator's, builds on the work instead of repeating it.

The value is a shared, durable record of what the project has actually learned. Glen is org-scoped, so the knowledge is shared across every researcher, every agent, and every tool in your group. A paper one collaborator's agent already evaluated is not re-surfaced as new, and because Glen is a standard MCP server, the same memory is readable from Claude Code, Cursor, your literature-search agent, or any other MCP client. The intellectual context that usually lives in scattered notes and one person's head becomes durable memory your agents draw on. You connect once over OAuth or an API key and let your research compound across sessions and teammates.

FAQ

Is Glen a reference manager or note-taking app?
No. Glen is shared memory for the AI agents you research with. It captures evolving judgments, ruled-out leads, methods, and conclusions, and feeds them back to your agents so they build on prior work automatically.
Can my whole research group share one memory?
Yes. Glen is org-scoped, so every collaborator's agent and every MCP client reads and writes the same memory of what the project has learned.