GetMyBot ナレッジベース

Knowledge sources

Give the AI your instructions, articles, and files, update them, and check what the agent finds for a customer's question.

このページの内容

To help an AI agent answer according to your rules, add materials such as delivery terms, course descriptions, instructions, or a price list. Each set of materials is a knowledge source. Find them under Knowledge base → Sources.

After you add a source, GetMyBot extracts its text and prepares it for search. This is called indexing. In Test retrieval, enter a customer's question and inspect the excerpts the agent would receive. For example, "Can I get a refund?" should find your refund terms.

You must also select the source for a specific agent under AI answers and agents. Merely uploading a file or creating an article does not give the agent access to it. Websites can refresh on a schedule; other source types refresh after a change or when you request it.

Source types

  • Knowledge base article: an existing article of this bot. Published articles become public sources, internal articles become agent-only sources, and drafts produce no searchable content at all.
  • Website: pages at allowed addresses that GetMyBot reads automatically.
  • File: a PDF, Word, Excel, PowerPoint, text, Markdown, or HTML upload.
  • Instruction: a short piece of policy text you write yourself, for example a refund rule the agent must always follow.
  • Fact: a single stored fact referenced by its identifier.

The source name is limited to 120 characters. The complete source configuration, including instruction text, is limited to 64 KB.

Visibility and priority

Every source has one visibility:

  • AI agent: only the agent can retrieve it.
  • Operators: reserved for operator-facing retrieval.
  • Public: may also appear in public surfaces such as the help center.

Files, instructions, and facts cannot be public. Only a published knowledge base article can have public visibility. Access is checked before search, so an agent restricted to public materials cannot receive an internal excerpt.

Priority ranges from −100 to 100. It is a tie-break: when two excerpts rank equally, the higher-priority source wins. It never overrides a clearly better text match.

Website sources

Enter the site addresses GetMyBot may read, for example https://example.com without a page path. An address with its protocol and domain is called an origin; you can add up to 20. Then add 1 to 100 pages where reading should begin. Only HTTPS is allowed.

Crawling stays inside those boundaries:

  • The site reader identifies itself as MyBotKnowledge/1.0 and checks each site's rules in robots.txt before fetching. It skips paths forbidden by Disallow and honors Crawl-delay for up to 10 seconds.
  • Only <a href> links that resolve back into the allowed origins are followed. Fragments are dropped and the utm_*, gclid, and fbclid parameters are removed before a URL is queued, so the same page is not indexed twice.
  • Before every request, the site address is resolved to an IP address again. Internal and service addresses (loopback, private, link-local, multicast, and cloud metadata) are blocked. Redirects follow the same rules, with a maximum of three.
  • A request times out after 15 seconds, the response body is capped at 8 MB, and only HTML, XHTML, plain text, and Markdown responses are accepted.
  • A single source indexes at most 100 pages.

The site reader does not sign in, submit forms, or load scripts, media, and embedded resources. It cannot read a page behind a login: publish the material or upload it as a file instead.

Refresh every (seconds) applies to website sources only. It must be at least 60 seconds and at most 366 days. Article, file, instruction, and fact sources are not re-fetched on a schedule; they are re-indexed when you change them or press Refresh.

File uploads

Uploads must already belong to this bot's media library and be at most 8 MB. The accepted types are PDF, DOCX, XLSX, PPTX, text/plain, text/markdown, text/html, and XHTML.

Extraction is deliberately narrow:

  • HTML keeps headings, lists, and link text; scripts, styles, navigation, headers, footers, asides, and templates are stripped.
  • PDFs are read as text only, up to 100 pages. Encrypted or malformed files are rejected instead of being partially parsed.
  • Office files are read from their document text parts only. Macros, embedded objects, and external relationships are ignored and never fetched. The archive must stay under 500 entries, a 100:1 compression ratio, 32 MB expanded, 200 000 XML nodes, and 500 000 characters of text.

Legacy binary .doc, .xls, and .ppt files are rejected with unsupported_type. Save them as DOCX, XLSX, or PPTX and upload again.

Indexing and statuses

A source moves through these states, shown on its card:

  • Waiting to index: created or changed, not processed yet.
  • Indexing: a worker holds the job.
  • Ready: the current content is searchable.
  • Needs refresh: the schedule is due or the content changed.
  • Failed: the last attempt ended with an error code.
  • Disabled: kept but excluded from every search.

To prepare a source for search, GetMyBot divides its text into fragments of about 1,500 characters with 200 characters of overlap. It then creates embeddings, numerical representations of meaning, in batches of 32. Only after all fragments are processed does the platform replace the old source version. Search therefore sees either the complete old or complete new version, never a partial update. If the content has not changed, processing is skipped.

The bot's AI integration creates embeddings using its base URL and key (Integrations). By default, text-embedding-3-small returns 1,536 numbers per fragment. Without a working AI integration, the source cannot become searchable.

One source can contain at most 100 documents, and one document at most 500 fragments. If processing stops unexpectedly, the job can pass to another worker after two minutes rather than blocking the source. The source card shows the number of attempts, processed documents and fragments, and documents with errors. Cancel indexing stops the current run.

What retrieval returns

Search considers both the words in a question and its meaning. For example, "How can I return a product?" can find an article about refund terms. It runs two searches:

  1. a lexical search over the PostgreSQL text index, and
  2. a semantic search over the vector index (cosine distance).

Both searches exclude other bots' content, sources hidden from this search, disabled sources, and deleted documents before ranking. Their result lists are then combined by position using reciprocal rank fusion. Source priority and indexing recency within 30 days break ties. Adjacent fragments from one document are merged.

The result is at most 12 excerpts of at most 800 characters each. A question is limited to 16 000 characters, and an agent may be restricted to at most 100 specific sources.

The agent stores the source, document, and fragment behind every excerpt it received, in rank order. An answer that cites anything else is rejected before it is sent, so a reply can never point to a draft, hidden, disabled, or deleted source.

Testing retrieval

Knowledge base → Test retrieval shows what the agent would find for your question without generating an answer. Each result shows its source, title, excerpt, match score, and reason ("Hybrid match"). Hidden internal details are not shown, and the number of tests in a short period is limited per account.

Use it before activating an agent: if the excerpts here are wrong, the answer will be wrong too.

When indexing fails

The last failure is stored as a stable code:

  • blocked_address: the URL, a redirect, or the resolved IP is outside the allowed origins or points at a private address.
  • robots_denied: robots.txt forbids that path.
  • unsupported_type: the content type or file format is not supported.
  • file_too_large: the download or upload exceeds the size limit.
  • expanded_too_large: the extracted text, page count, or archive contents exceed the limits above.
  • malformed_file: the file or response could not be parsed.
  • fetch_timeout: the site did not answer in time.
  • rate_limited: the embedding provider throttled the request; the job is retried.

Fix the cause and press Refresh. Errors never include the fetched text, so they are safe to share with support.

Deleting, disabling, and retention

Disable keeps the configuration but removes the source from every search; Enable puts it back in the queue. Delete removes the source together with its documents and fragments.

A source that is selected by an active agent cannot be deleted: the request fails with a conflict. Deselect it in the agent first, or pause the agent.

Isolation

Sources, documents, fragments, and processing jobs are tied to a bot. Search is restricted to that bot's content before retrieving results, so another account's content cannot appear even if the text is identical.

API and MCP

The REST API uses the knowledge:read and knowledge:write scopes:

  • GET/POST /api/bots/{botID}/knowledge/sources
  • GET/PATCH/DELETE /api/bots/{botID}/knowledge/sources/{sourceID}
  • POST /api/bots/{botID}/knowledge/sources/{sourceID}/upload
  • POST /api/bots/{botID}/knowledge/sources/{sourceID}/refresh
  • POST /api/bots/{botID}/knowledge/sources/{sourceID}/cancel
  • POST /api/bots/{botID}/knowledge/sources/{sourceID}/disable
  • POST /api/bots/{botID}/knowledge/sources/{sourceID}/enable
  • GET /api/bots/{botID}/knowledge/sources/{sourceID}/index-status
  • GET /api/bots/{botID}/knowledge/sources/{sourceID}/documents
  • POST /api/bots/{botID}/knowledge/retrieval-test

MCP exposes read-only tools: list_knowledge_sources, get_knowledge_source_status, and test_knowledge_retrieval. There is no MCP tool that changes a source or downloads raw hidden documents.