DACIX — The Store for AI Agents (MCP Server)

Remote MCP server for dacix.store — a machine-first marketplace where AI agents buy agent templates and pay-per-call data services, with Stripe checkout and automatic fulfillment.

• Endpoint: https://dacix.store/mcp (streamable HTTP, no auth needed to connect)
• Registry: published in the official MCP registry as store.dacix/store
• Docs: OpenAPI · llms.txt · agents.json

Configuration

Add DACIX to any MCP client that supports streamable HTTP (Claude Desktop, Cursor, Windsurf, ...):

{
  "mcpServers": {
    "dacix": {
      "type": "http",
      "url": "https://dacix.store/mcp"
    }
  }
}

Claude Code one-liner:

claude mcp add --transport http dacix https://dacix.store/mcp

No connection parameters needed — your agent creates its own account with the store_register tool and receives an API key plus 25 free credits.

Tools

• store_catalog: List everything for sale at dacix.store — agent templates (portable JSON agent definitions) and API credit packs. No auth needed.
• store_register: Create a DACIX account and get an API key (dacix_sk_...) plus 25 free trial credits.
• store_account: Check your account — credit balance and owned templates.
• store_buy: Buy a product — returns a Stripe checkout URL; fulfillment is automatic after payment.
• store_order_status: Check an order's status after payment (pending → paid).
• store_download_template: Download a purchased agent template (full JSON: system prompt, tools, schemas).
• web_crawl: Fetch any public web page server-side and get clean readable text back. 5 credits.
• web_search: Search the web via DuckDuckGo — structured results with titles, URLs, and snippets. 5 credits.
• web_fetch: Fetch a page with enhanced text extraction — scripts/styles/nav stripped, clean readable text + word count. 5 credits.
• analyze_seo: Full SEO audit of a page — title, meta description, Open Graph tags, heading structure, link counts, word count. 5 credits.
• compare_pages: Fetch 2–5 pages and compare titles, word counts, and content previews side by side. 10 credits.
• ro_company_lookup: Romanian company registry profile by CUI (tax ID) — legal name, active/struck-off status, VAT registration, CAEN, address. ~2M companies. 10 credits.
• ro_company_search: Search Romanian companies by name — returns CUI, legal form, county, status per match. 10 credits.
• ro_company_search_by_address: Find every Romanian company registered at an address (due diligence: who is registered here?). 10 credits.
• ro_company_financials: Yearly financials for a Romanian company by CUI — net turnover, profit/loss, employees, equity, assets, debt, multi-year history. 20 credits.
• ro_company_list: Browse Romanian companies with filters — county, active vs struck-off, VAT-registered. Paginated. 10 credits/page.
• classified_categories: Category tree of micapublicitate.online (Romanian classifieds) — pick a category before posting an ad. Free.
• post_classified_ad: Publish a real classified ad on micapublicitate.online (90-day listing) — returns the public URL. 20 credits.
• city_info: Practical civic info for 28,674 cities worldwide as clean markdown — parking, events, public transport, city hall, laws, quiet hours and ~30 more topics. 5 credits.

Failed lookups (not found / upstream errors) are never charged.

What's for sale

• Agent templates ($19–$29 one-time): portable JSON agent definitions — production-grade system prompt, tool schemas, strict output schema, config. Load into the Claude API, LangChain, or any runtime. Web research, invoice extraction, Romanian legal research, SEO content, support triage.
• Credit packs ($9 / $49 / $299): fuel fo

... View full README on GitHub