علّم مساعدك محتواك الخاص

Busymate AI makes your mate answer from your own content: point it at a website you run, or paste in text you wrote, then publish. Every answer is sourced, and your mate admits when it cannot be sure.

1. Add a website source

Point the assistant at your help site. The crawler reads the site's own sitemap first (the Sitemap: lines in robots.txt, or /sitemap.xml), then follows links for anything the sitemap misses. It is robots-aware, stays on the same origin, and runs asynchronously. A source for a section, such as https://example.com/docs, reads only that section's sitemap entries.

How many pages it reads is set by your plan, across all your sources: Free 200, Starter 2,000, Growth 20,000, Scale 100,000. A new source reads the whole site up to that allowance. You can set a lower page limit per source.

If your site publishes its own MCP tools, your mate uses those tools instead of crawling it.

• Console: Console → Knowledge → Add source → URL. Leave the page limit empty to read the whole site.
• MCP: add_tenant_knowledge_source with url (and optionally max_pages); then list_tenant_knowledge_sources, reindex_tenant_knowledge_source, set_tenant_knowledge_source_enabled and delete_tenant_knowledge_source to manage it.

A source shows its status, the pages and chunks indexed, the last index time and the last error, so an empty crawl is visible rather than silent.

If the site publishes its own MCP tools, it is not crawled: your mate answers from those tools and cites their pages. See Make your site's knowledge available with MCP.

Coverage: what was read and what was not

Console → Index has a Coverage card, and MCP get_index_status returns the same coverage block:

• your plan's page allowance, how much of it is used and how much is left;
• per source: whether pages were found through the sitemap or by following links, how many pages the sitemap lists, and how many of them are not read because of the plan or the source's page limit;
• URLs the sitemap lists that are files (PDFs, images), outside the source's section, or blocked by robots.txt;
• pages that were skipped, failed or excluded, grouped by reason (for example "robots.txt does not allow it", "no readable text on the page", "the site answered HTTP 404").

Re-crawl schedule

Each source is re-read on its own schedule: every 6 or 12 hours, every day (the default), every 3 days or every week. Pages that keep changing are re-read sooner, and pages that never change are re-read less often. Set it in Console → Index → Re-read, or with MCP set_index_config (sync_interval_minutes, 60 to 10,080). recrawl_index re-reads a whole source or one page now.

2. Or publish curated text

Publish text you wrote — facts, how-tos, references, glossary entries — as knowledge_sources on the revision itself (publish_tenant_runtime). Each source has a caller-stable key, a label, a kind (fact, howto, reference or glossary) and its content.

• Limits: at most 20,000 characters per source, 40 sources, 40,000 characters in total. An over-limit payload is refused before anything is written.
• Re-publishing the same source replaces it.

The Shopify app uses exactly this path when it learns your products, policies and pages.

Pinned answers: the answers you want given word for word

For the questions where the wording matters (refunds, pricing, opening hours, legal), pin your answer. When a customer asks that question, your mate answers with your answer, ahead of every website and document source.

• Console: Console → Knowledge → Pinned answers. Write the question, other ways customers ask it (up to 10), and the answer (up to 4,000 characters). You can start one from a canned response.
• MCP: list_tenant_pinned_answers, upsert_tenant_pinned_answer, delete_tenant_pinned_answer, and test_tenant_pinned_answer to see which pinned answer a question would get.

The same wording, or a question that is very close to it, is answered with your text as written (in the customer's language if they write in another one). A question that is only similar gets your answer when it actually answers what was asked. Add other phrasings to catch more ways of asking. Pinned answers are live as soon as you save them, with no publish step, and each one shows how often it was used.

3. Publish to go live

Content reaches your mate through a published version: select the sources, run the checks, publish. Published versions are frozen, so what your customers see only ever moves forward to a complete setup. See Getting started.

4. Test what it finds

• Ask your mate a question only your content can answer. The reply cites the source.
• Over MCP, search_tenant_knowledge returns the matching excerpt with where it came from, so you can check what the assistant would read before a customer does.
• Ask something your content does not cover. Your mate says it is not sure and, if handoff is on, offers a person.

Starter suggestions

The welcome screen's suggestion pills are generated from your published content. Publish good sources and the first prompts your customers see are already about your product.

The wording rule

Say it the way it works: the assistant answers only from your content, with sources, and admits when it cannot be sure. Claim nothing beyond that sentence.

Manage from any MCP client

Every step above is available from any MCP client connected with your own account:

claude mcp add --transport http busymate-ai https://busymate.ai/mcp

Technical details

For developers: curated text is stored in 2,000-character parts, and every answer drawn from a source carries a [K:xxxx] citation the reader can follow back to the exact part.

Verify

1. A content-only question returns a cited answer.
2. An out-of-scope question is declined; with handoff on, a person is offered.
3. Change a page, reindex the source, ask again — the answer follows the change.

Next steps

• Make your site's knowledge available with MCP — publish docs tools instead of being crawled.
• Governance and models — the features and limits around the assistant.
• AI customer support with human handoff — grounded answers plus a person.
• AI sales and onboarding assistant — pre-sales answers from your own docs.