Ppl

io.github.ppl-crmv2.0.0Updated Oct 5, 2026

ppl: the personal CRM for AI agents. Structured relationship memory over hosted MCP.

VerifiedStreamable HTTPWeb executableProductivity & Workflow Knowledge & Memory

Overview

AI-generated overview

Gives an assistant a hosted personal CRM to read and write relationships, notes, journal entries, tasks, and reminders.

What it does: ppl is a hosted personal CRM exposed as an MCP server. It stores relationships, notes, journal entries, tasks, and reminders, and lets an assistant read that context before acting and write back to it afterwards. The README describes connecting over Streamable HTTP with a Bearer token, plus an agent connection flow that requests approval and returns an API token. It is hosted-only, with no downloadable server.
When to use it: Use it when you want an assistant to keep durable relationship memory across sessions, such as recalling who someone is, what was discussed, or what follow-ups are pending. It suits personal CRM and journaling workflows rather than general data storage.
Requirements: A remote MCP endpoint reached over Streamable HTTP at the provider's hosted address. A Bearer token is required; the README describes obtaining one through an agent connection request that a human approves, which may create an account inline. No local runtime or package is needed.

Before you install

The server is hosted and holds personal relationship data, notes, and journal content, so anything written is stored with the provider. Access depends on a Bearer token, which should be treated as a secret. The README states the assistant writes to the CRM after acting, so review what gets recorded. No payment or wallet details appear in the input.

Installation

In SourceWeft

Open Ppl in the dashboard and add it to a workspace.
Enable the server for the chats that should use its tools.

Web executable via Streamable HTTP. Remote servers run from the web runtime once configured in a workspace.

Other MCP clients

Add this to your client's mcpServers config.

{
  "mcpServers": {
    "ppl": {
      "type": "http",
      "url": "https://withppl.com/mcp"
    }
  }
}

README

ppl

Relational memory for you and your AI.

ppl is the personal CRM for humans with AI agents. Your relationships, notes, journal, tasks, and reminders live here. Your AI reads from it before acting, and writes to it after acting.

Connect your AI

Hosted MCP server: https://withppl.com/mcp

ppl is hosted-only. There is no downloadable server. Your agent connects over Streamable HTTP with a Bearer token.

One-click setup

Send your human to withppl.com/agents. If they use Cursor and are logged in, it's one click. New users get an account created inline, no signup form first.

Agent connection flow

POST https://withppl.com/api/agent/connect/request with {"client_name": "Your agent"}
Show the human the approve_url. They approve once (creating an account inline if needed).
Poll the poll_url until approved and save the api_token.
Use the token with the MCP server or REST API.

Full machine-readable spec: withppl.com/.well-known/ppl.json

MCP registry

Registered as io.github.ppl-crm/ppl in the official MCP registry. See server.json.

Examples

Python MCP client - connect with the MCP Python SDK
Node MCP client - connect with the MCP TypeScript SDK
ppl-memory PyPI package - LangGraph, CrewAI, and AutoGen adapters

Docs

Agent integration guide - session prescription, tools, and patterns
OAuth for MCP clients - Dynamic Client Registration flow

License

MIT. See LICENSE.

Source: README.md at commit 8c9b1ed

Tools

Tool metadata has not been indexed yet.

Version history

v2.0.0LatestOct 5, 2026

Ppl