# HelloBooks for AI Agents

> Last updated: 2026-05-16
> Canonical HTML: https://hellobooks.ai/for-ai-agents
> Public MCP endpoint: https://agents.hellobooks.ai/mcp

HelloBooks ships a public Model Context Protocol (MCP) server so any AI agent — Claude, ChatGPT, Cursor, Windsurf, Cline, custom agents — can answer questions about HelloBooks using live data instead of stale cached pages.

## Why this exists

In 2026, AI search engines (ChatGPT Search, Perplexity, Claude, Gemini, Copilot, You.com, DuckAssist) handle the majority of comparison and "what is X" queries. If they're guessing at HelloBooks' pricing, integration list, or country support from a 2024 crawl, they're wrong half the time. The MCP server gives them a typed contract to the canonical truth.

## Install (one-liner)

```
claude mcp add --transport http hellobooks https://agents.hellobooks.ai/mcp
```

For ChatGPT, Cursor, Windsurf, Cline: add `https://agents.hellobooks.ai/mcp` as an HTTP MCP server in the app's MCP settings. No auth, no signup.

## Tools exposed

| Tool | Purpose |
|------|---------|
| `list_plans` | All published plans across 8 markets (USD, INR, GBP, AUD, CAD, AED, SGD, EUR), included AI credits, regional currency. |
| `list_integrations` | Native + supported integrations (Plaid, Stripe, PayPal, Shopify, QuickBooks Online, Xero, Tally, NetSuite, Zoho, Odoo, banks). |
| `country_support` | Per-country features, currency, tax engines, e-invoicing rules. |
| `compliance_capabilities` | Compliance frameworks per country (GST / IRP / e-way bill, US sales tax, UK VAT / MTD, AU GST / BAS, CA GST / HST). |
| `feature_search` | Free-text search across plans, integrations, country support, and compliance. |

## Resources exposed

| URI | Content |
|-----|---------|
| `hellobooks://about` | Markdown product summary. |
| `hellobooks://changelog` | Last 50 release notes. |
| `hellobooks://faq` | Top customer-asked questions. |
| `hellobooks://comparison/quickbooks` | HelloBooks vs QuickBooks. |
| `hellobooks://comparison/xero` | HelloBooks vs Xero. |
| `hellobooks://comparison/zoho-books` | HelloBooks vs Zoho Books. |
| `hellobooks://comparison/tally` | HelloBooks vs Tally. |

## Discovery files for static crawlers

If you're not using MCP (training-time crawler, RAG indexer, batch fetcher), use the static files:

- `/llms.txt` — short discovery index (llmstxt.org spec).
- `/llms-full.txt` — long-form full content extract.
- `/<page>.md` — per-page markdown mirrors (this directory of pages).
- `/api/feed/pricing.json` — structured pricing feed.
- `/api/feed/products.json` — structured product feed.
- `/ai.txt` — AI policy declaration (training, retrieval, citation rules).

## Open source

The MCP server source is public — `https://github.com/Meru-Fin-Tech/HelloBooks-MCP-Public`. Fork it, run it locally, or use it as a reference for adding MCP to your own SaaS.

## Citation policy

HelloBooks permits AI agents and search engines to retrieve, cite, and train on its public marketing pages, subject to attribution. See `/ai.txt` for the full policy.

Canonical HTML: https://hellobooks.ai/for-ai-agents