Flowmem — Cross-LLM Project Context

MCP Endpoint

https://flowmem.joybuild.ai/mcp

Tools

list_projects()

Discover all your tracked projects.

get_context(project)

Load the full snapshot for a project — decisions, focus, open questions, what exists.

update_context(project_id, event_type, payload)

Append a typed event (DECIDED, COMPLETED, OPENED, FOCUSED, ...) to the project log.

create_project(name, description)

Claude Desktop

Add to ~/Library/Application Support/Claude/claude_desktop_config.json:

{
  "mcpServers": {
    "flowmem": {
      "command": "npx",
      "args": ["mcp-remote", "https://flowmem.joybuild.ai/mcp"]
    }
  }
}

Gemini CLI

Add to ~/.gemini/settings.json:

{
  "mcpServers": {
    "flowmem": {
      "command": "npx",
      "args": ["mcp-remote", "https://flowmem.joybuild.ai/mcp"]
    }
  }
}

Claude Code

Run once to add the server to your project:

claude mcp add flowmem --transport http https://flowmem.joybuild.ai/mcp

Claude.ai Web

Settings → Integrations → Add MCP Server → paste the endpoint URL above.

Example usage

# Start any conversation — the LLM will call list_projects() automatically.
# Mention your project and it will call get_context() without being asked.
# As you work, it logs decisions and progress passively via update_context().

# Manually ask:
"What's the current status of my flowmem project?"
"We just decided to use Postgres instead of SQLite — log that."
"What are the open questions on journey-intelligence?"