---
title: "Update knowledge base settings"
method: PATCH
path: "/kb"
tags: ["Knowledge Base"]
---

# Update knowledge base settings

`PATCH /kb`

## Request

- Base URL: `https://api.weav.com/v1`
- URL: `https://api.weav.com/v1/kb`
- Auth: HTTP bearer

## Request body

- object
  - `improvement_enabled` boolean: Whether teammate corrections can propose KB drafts.
  - `mcp_enabled` boolean: Whether the public MCP server and WebMCP tools are available at the KB's /mcp URL.
  - `social_links` object: Partial HTTP(S) URLs, merged with stored links; null or empty removes a link.
    - `x` string, nullable
    - `linkedin` string, nullable
    - `facebook` string, nullable
    - `instagram` string, nullable
    - `youtube` string, nullable
    - `github` string, nullable
    - `discord` string, nullable
    - `website` string, nullable
  - `banner` object: Partial banner update. Enabling a banner requires nonblank content.
    - `enabled` boolean: Cannot be true unless content is nonblank.
    - `content` string, nullable: One line of inline Markdown (links, bold, italic, and inline code). Leading/trailing whitespace is trimmed; CR and LF are rejected.
    - `dismissible` boolean
    - `type` 'info' | 'warning' | 'critical': Default background colors are #2052EB for info, #B45309 for warning, and #B91C1C for critical.
    - `light_color` string, nullable: Optional 6-digit hex background override for light mode.
    - `dark_color` string, nullable: Optional 6-digit hex background override for dark mode.
  - `seo` object: Partial SEO settings update; values omitted from nested objects are preserved.
    - `indexing_enabled` boolean
    - `description` string, nullable: One line of plain text. HTML is stripped; ampersands are preserved.
    - `agent_instructions` string, nullable: Instructions for AI agents; HTML is stripped and line breaks are preserved.
    - `og_image_url` string, uri, nullable: HTTPS URL for the organization's default social preview image.
    - `verification` object: Provider code or a full pasted provider meta tag.
      - `google` string, nullable
      - `bing` string, nullable
      - `yandex` string, nullable
    - `content_signals` object
      - `ai_train` boolean
      - `search` boolean
      - `ai_input` boolean

## Response `200`

Success

- object
  - `seo` object
    - `indexing_enabled` boolean
    - `description` string, nullable
    - `agent_instructions` string, nullable: Instructions for AI agents; line breaks are preserved.
    - `og_image_url` string, uri, nullable
    - `verification` object
      - `google` string, nullable
      - `bing` string, nullable
      - `yandex` string, nullable
    - `content_signals` object
      - `ai_train` boolean
      - `search` boolean
      - `ai_input` boolean
  - `sitemap_url` string, uri

## Other responses

- `422`: Validation error

---

[API](https://skmtc.dev/weav/apis/weav-api.md) · [All operations](https://skmtc.dev/weav/apis/weav-api/llms.txt) · [OpenAPI document](https://skmtc.dev/weav/apis/weav-api/revisions/15970f126c0e?raw)