DEVELOPER GATEWAY
Deploy design intelligence directly into your production workspaces. Choose the visual workspace playground, program against raw API endpoints, or spin up the Model Context Protocol (MCP) server inside your local AI environment.
NEW TO THE PLATFORM? START HERE
Follow our 5-Minute Quickstart guide to link Claude to your custom API key. Run your first autonomous design system scrape in minutes.
CHOOSE YOUR PATHWAY
Visual Playground
Test scrapes visually from your browser. Input URLs, run synchronous mappers, review extracted variables, and export JSON config packages in an interactive UI.
Developer API
Integrate raw endpoints directly into your scripts or production backends. Supports custom modes, concurrency controllers, and robust idempotency retry hooks.
Claude MCP Server
Equip your LLM agents with design capability. Our server lets AI models analyze layouts, query color tokens, and generate Figma component codes locally.
PLAN CAPABILITIES & QUOTAS
| Feature / Quota | Free Tier | Starter ($4/mo) | Pro ($13/mo) | Enterprise ($29/mo) |
|---|---|---|---|---|
| Credit Quota / Month | 10 cr (Test) | 50 cr | 200 cr | 500 cr |
| Allowed Modes | Basic | Basic, Enhanced, Deep | Basic, Enhanced, Deep | Basic, Enhanced, Deep |
| Deduction Cost (cr) | Basic: 1 | Basic: 1 / Enhanced: 2 / Deep: 5 | Basic: 1 / Enhanced: 2 / Deep: 5 | Basic: 1 / Enhanced: 2 / Deep: 5 |
| Concurrent Deep Jobs | Locked (0) | 1 | 2 | 3 |
| Max Queue Depth | Locked (0) | 2 | 10 | 25 |
| Max Pages per Job | 1 | 3 | 10 | 25 |
| API Scopes | `scrape` | `scrape`, `enrich` | `scrape`, `enrich`, `generate` | `scrape`, `enrich`, `generate` |
HUB FREQUENTLY ASKED QUESTIONS
Q: How do I choose between visual scraping, API endpoints, and the MCP server?
Use the Visual Playground for one-off manual audits and immediate ZIP components exports. Automate with the API Reference when linking the scraper into database records or custom scripts. Run the MCP Server to equip AI coding assistants (like Claude) with native visual parsing capabilities.
Q: How are credit balances billed?
Credits are pre-deducted at the start of a request. Successful scrapes charge 1 credit (Basic), 2 (Enhanced), or 5 (Deep). If an asynchronous scraper worker fails due to an internal system error, the balance is automatically refunded to your key within 3 minutes.
Q: What data does the system retain?
No sessions, cookies, or credentials from your target websites are cached. Extracted packages are stored locally in our pruning directory and deleted permanently 1 hour after generation. Check our dedicated Security & Trust page for more verification.