sleds — Shared AI memory for teams

Claude

MCP Native

Full bidirectional context via Model Context Protocol. Works with Claude Desktop and claude.ai Projects. Read threads, write observations, search context, share assets — all natively.

Read/write threads & observations

Semantic search across all context

Share and retrieve assets

Real-time sync with other tools

Works with Desktop + Web

Quick Setup

// claude_desktop_config.json{
  "mcpServers": {
    "sleds": {
      "command": "npx",
      "args": ["sleds-mcp"]
    }
  }
}

Cursor

MCP Native

Your coding AI gets full project context. Architecture decisions, API specs, design tokens — Cursor sees everything the team has decided, without you copy-pasting a thing.

Auto-loads relevant specs while coding

Writes observations as you build

References shared assets inline

Sees decisions from other tools

MCP config in Settings or .cursor/mcp.json

Quick Setup

// .cursor/mcp.json{
  "mcpServers": {
    "sleds": {
      "url": "https://api.sleds.ai/mcp"
    }
  }
}

ChatGPT

Custom GPT Actions

Build a Custom GPT that's aware of your entire project. Connect via OpenAPI Actions spec — ChatGPT gets full CRUD access to your sled's threads, assets, and search.

OpenAPI 3.0 Actions integration

Full CRUD on threads & assets

Semantic search across context

Custom GPT system prompt included

Works with any GPT-4/4o model

Quick Setup

// Import as Custom GPT Action1. Create a new Custom GPT
2. Import OpenAPI spec from:
   api.sleds.ai/openapi.json
3. Add your API key
4. Start chatting with context

Gemini / AI Studio

MCP + REST API

Google's AI tools connect through MCP for native integration or REST API for AI Studio function calling. Full access to shared context either way.

MCP for native tool integration

REST API for AI Studio functions

Read/write all context types

Search and link capabilities

Structured data support

Quick Setup

// AI Studio Functionconst sleds = await fetch(
  "api.sleds.ai/api/spaces/
  my-sled/search"
, { headers: { Authorization }
});

Figma

Plugin

Push design tokens, component specs, and visual decisions from Figma directly into your sled. Engineering AI tools reference them instantly — no handoff docs needed.

Push design tokens to sled

Share component specifications

Reference decisions from Figma

Design → code context bridge

Coming: auto-sync on publish

Quick Setup

// Install Figma Plugin1. Search "sleds" in Figma
   Community plugins
2. Install and authenticate
3. Push tokens with one click
   Coming Q2 2026

Slack

Bolt SDK

Bidirectional Slack integration. Capture context from conversations with emoji reactions or slash commands. Get notified in channels when your sled updates.

🧠 emoji reaction to capture

/sleds slash command to save context

/sleds search from any channel

Channel notifications on updates

Configurable per-sled channels

Quick Setup

// Install Slack App1. Add sleds to your workspace
2. Connect to a sled in Settings
3. React with 🧠 to capture
4. Use /sleds to search
   Available now ✓

GitHub

GitHub App

Connect your repos to sleds via GitHub App. PR events, issues, and deployments flow into sled threads automatically. Frost can create issues and comment on PRs.

Org-level webhook events

PR & issue sync to threads

Deploy status tracking

Frost can create issues & comment

Coming Q2 2026

Quick Setup

// Install GitHub App1. Install sleds GitHub App
   on your org or repo
2. PR, issue & deploy events
   sync to sled threads
3. Frost comments on PRs
   Coming Q2 2026

Linear

REST API

Sync project tracking with context. Issues and project updates from Linear flow into your sled, giving AI tools awareness of what's being built and when.

Issue sync to sled threads

Project status tracking

Cycle and milestone awareness

Frost can create/update issues

Coming Q2 2026

Quick Setup

// Connect Linear workspace1. Authorize sleds in Linear
2. Select projects to sync
3. Issues auto-create threads
4. Cycle status updates Frost
   Coming Q2 2026

Connect everything

Ready to ride?