Save and recall project context across AI tools
A persistent, shared markdown memory that allows different AI assistants to read and write the same context, eliminating the need to re-paste information when switching between tools like Claude, Cursor, or ChatGPT.
How it works
The agent connects to the Agentage Memory MCP server. It writes user-defined notes, decisions, or architectural constraints as plain markdown files with YAML frontmatter to the shared memory store. When the user switches to another AI tool, the agent can read these files to recall previous context, ensuring continuity. The user should verify that the memory is correctly formatted and that the specific note or decision is accessible via the lexical search or file listing before relying on it for complex tasks.
Prompt
Paste this into an agent that has the servers below connected. Adjust the names and dates to your own.
Save the following project decision to my shared memory: 'We chose SQLite for the local cache because of its zero-config nature and ACID compliance.' Then, recall any previous notes regarding database selection to ensure consistency.MCP servers it needs
Source
The vendor page this playbook is drawn from, and the passage that supports it.
Ask it to save or recall a note - every AI you connect shares the same memory.Shared markdown memory for Claude, ChatGPT & Cursor | Agentage