Skip to main content
Workweaver Workweaver / Developers
Home Documentation Agent Knowledge

Developer Documentation

Connect your tools to Workweaver. CLI commands, MCP integration, SDK libraries, public API endpoints, and browser extension setup.

Workweaver CLI

The Workweaver CLI provides command-line access to WorkMemory, mission control, and tenant management. Install via pip and authenticate with your Workweaver account.

Installation

pip install workweaver

Authentication

ww login

Connect WorkMemory

ww memory bootstrap --harness cursor

Available Commands

ww login

Authenticate with your Workweaver account

ww connect

Connect tools and services to WorkMemory

ww memory

Query and manage WorkMemory contents

ww run

Execute workflows and missions

Help

Run ww --help for a complete list of commands and options.

WorkMemory MCP Server

The Model Context Protocol (MCP) server provides WorkMemory integration for AI assistants. Connect Claude Code, Cursor, OpenCode, and other MCP-compatible tools.

Publishing soon

The standalone npm package and copy-paste JSON snippets are not publicly installable yet. Today, supported harnesses should use the Workweaver CLI bootstrap path, and everyone else should treat the package name as release-planning metadata until publication lands.

Supported harness bootstrap

ww memory bootstrap --harness cursor

Manual MCP configuration

Public manual JSON snippets will ship with the standalone package release. Until then, use ww memory bootstrap --harness cursor (or the equivalent supported harness value) so Workweaver writes the correct local MCP configuration for you.

Available Tools

memory_remember

Save a fact, preference, decision, or observation to memory

memory_recall

Find relevant context from memory — ask naturally

memory_forget

Remove outdated or incorrect information from memory

memory_status

Check WorkMemory health, storage usage, and memory count

Documentation

See the live MCP tools catalog for the currently shipped public tool surface.

Workweaver SDKs

Official SDK libraries for TypeScript, Python, LangChain, OpenAI, and Vercel. Integrate WorkMemory and Workweaver execution into your applications.

Publishing soon

The SDK distribution names are fixed, but the public npm and PyPI releases are not live yet. We will replace these placeholders with real install commands only after each package exits 0 on a clean machine.

TypeScript SDK

@workweaver/memory-client

Full-featured TypeScript SDK with type definitions. Public npm release is publishing soon.

Python SDK

workweaver-memory

Python SDK with async support and pydantic models. Public PyPI release is publishing soon.

LangChain

workmemory-adapter-langchain

LangChain integration for AI agents. Public PyPI release is publishing soon.

OpenAI

workmemory-adapter-openai

OpenAI integration for shared WorkMemory flows. Public PyPI release is publishing soon.

Vercel AI SDK

@workweaver/workmemory-adapter-vercel

Vercel AI SDK integration. Public npm release is publishing soon.

Public API

RESTful API for WorkMemory, mission execution, and tenant management. Authenticate with Bearer tokens (format: ww_sk_*).

Base URL

https://api.workweaver.ai/api/v1/public

Authentication

curl -H "Authorization: Bearer ww_sk_your_token_here" \
  https://api.workweaver.ai/api/v1/public/memory/search

Memory Endpoints

GET /memory/search

Search WorkMemory by semantic similarity

POST /memory/add

Add facts and episodes to WorkMemory

GET /memory/recall

Recall recent memories and patterns

DELETE /memory/{id}

Delete specific memory by ID

Execution Endpoints

POST /execute

Execute a workflow or mission

GET /status/{run_id}

Check execution status and results

Example: Search Memory

curl -X POST "https://api.workweaver.ai/api/v1/public/memory/search" \
  -H "Authorization: Bearer ww_sk_your_token_here" \
  -H "Content-Type: application/json" \
  -d '{
    "query": "Acme Corp preferred meeting times",
    "limit": 5
  }'

Documentation

See the main documentation for detailed API reference and examples.

Browser Extension

Workweaver browser extension for ChatGPT, Claude, and Gemini. Provides WorkMemory context and execution capabilities directly in your AI chat.

Installation

The browser extension is available for Chrome, Brave, and other Chromium-based browsers.

Features

ChatGPT Integration

Inject WorkMemory context into ChatGPT conversations

Claude Integration

Enhanced Claude conversations with WorkMemory

Gemini Integration

Google Gemini with WorkMemory context

Privacy First

Local processing with encrypted sync

Setup

  1. Install the extension from the Chrome Web Store
  2. Click the extension icon in your browser toolbar
  3. Sign in with your Workweaver account
  4. Enable integrations for ChatGPT, Claude, or Gemini
  5. Start chatting with AI assistants that have WorkMemory context

Permissions

The extension requires:

  • Read access to active chat pages
  • Storage permission for local caching
  • Network permission for Workweaver API calls

Documentation

The setup steps on this page are the current public source of truth. Longer-form browser-extension documentation will publish alongside the storefront listing.