mintlify · mintlify · Jun 30, 2026
diff --git a/ai/model-context-protocol.mdx b/ai/model-context-protocol.mdx
@@ -14,7 +14,7 @@
 Your search MCP server exposes tools for AI applications to search and retrieve your content. Your users must connect your search MCP server to their tools.

 <Tip>
  Looking to let agents edit your content instead of read it? Use the [admin MCP server](/ai/mintlify-mcp) for an authenticated MCP server that exposes branching, page editing, navigation, and `docs.json` tools to trusted agents.
 </Tip>

 ### How MCP servers work
@@ -26,21 +26,22 @@
 - Each tool call happens during the generation process, so the AI application uses up-to-date information from your site to generate its response.

 <Tip>
  Some AI tools like Claude support both MCP and skills. MCP gives access to your content, while skills instruct agents how to use that content effectively. They're complementary and connecting your MCP server gives agents access to both.
 </Tip>
 
 ### MCP tools
 
-Your search MCP server provides two tools that agents can use:
+Your search MCP server provides three tools that agents can use:
 
 - **Search**: Searches across your site to find relevant content, returning snippets with titles and links. Use this to discover information or find pages matching a query.
 - **Query docs filesystem**: Reads and navigates your site's virtual filesystem using shell-style commands. Use this to browse and retrieve content, or extract specific sections—including batch reads across multiple pages in a single call.
+- **Submit feedback**: Reports a documentation issue back to your team when an agent finds a page that is incorrect, outdated, confusing, or incomplete. The agent supplies the page path and a description of the problem, which is recorded as unhelpful feedback against that page so your team can fix it. This is the only tool that writes data; the rest of the server is read-only.
 
-Agents determine when to use each tool based on the context of the conversation. For example, an agent might first search your site to find relevant pages, then use the query docs filesystem tool to read the full content of the most relevant results.
+Agents determine when to use each tool based on the context of the conversation. For example, an agent might first search your site to find relevant pages, then use the query docs filesystem tool to read the full content of the most relevant results, and call submit feedback if it discovers a problem with the page along the way.
 
 ### MCP resources
 
 Your search MCP server also exposes your [skill.md files](/ai/skillmd) as MCP resources. Agents connected to your search MCP server can discover and access your skill files without installing them separately.

 `Skill.md` resources appear in the search MCP server's resource list and contain the capability descriptions Mintlify generates or that you define in your [custom skill files](/ai/skillmd#custom-skill-files).

@@ -69,17 +70,17 @@
 * For content with partial authentication, where some pages are public and others require login, you must enable your search MCP server before users can access it. Unauthenticated users can search public content. Users who authenticate can search all content they have permission to access based on their [user groups](/deploy/authentication-setup).
 * For content where all pages require authentication, you must enable your search MCP server before it is available to users. Users must authenticate before connecting to your search MCP server. Your search MCP server searches only the content each user has access to based on their [user groups](/deploy/authentication-setup).

 If authentication is enabled, configure your hosted search MCP server from **Settings → Security & access → MCP** in your dashboard. The authenticated search MCP server is available at the `/authed/mcp` path of your site URL.

 <Note>
  The **MCP** settings page only appears when you enable [authentication](/deploy/authentication-setup) for your site. If your site doesn't require authentication, your search MCP server is still accessible at the `/mcp` path of your site URL.
 </Note>

 The `/mcp` and `/authed/mcp` paths are reserved for search MCP servers and cannot be used for other navigation elements.

 ### Discovery endpoint

 Mintlify hosts a discovery document at `/.well-known/mcp` for agents and tools to locate your search MCP server without prior configuration.

 `GET /.well-known/mcp` returns a JSON document describing your search MCP server:

@@ -165,7 +166,7 @@
 | `tools[].inputSchema` | JSON Schema for the tool's input parameters. |
 | `tools[].annotations` | Optional [MCP tool annotations](https://modelcontextprotocol.io/specification/2025-06-18/server/tools#tool-annotations) that describe tool behavior. |

 The built-in search and query docs filesystem tools are read-only and non-destructive with the following annotations:

 - `readOnlyHint: true`—the tool does not modify any state.
 - `destructiveHint: false`—the tool has no destructive side effects.
@@ -187,7 +188,7 @@

 By default, your MCP server is only available for localhost tools. To allow web-based tools to connect, add the redirect domains for the AI tools. A redirect domain is the hostname that an AI tool uses after a user completes authentication like `claude.ai` or `app.cursor.ai`. Your users' AI tools cannot complete authentication unless their redirect domain is on this list.

 Your hosted MCP server is enabled automatically when your site has [authentication](/deploy/authentication-setup) configured. To allow AI tools to complete authentication, add their redirect domains in your dashboard.

 <Steps>
  <Step title="Open the hosted MCP settings">

diff --git a/es/ai/model-context-protocol.mdx b/es/ai/model-context-protocol.mdx
@@ -37,12 +37,13 @@ Cuando una aplicación de IA se conecta a tu servidor Search MCP, puede buscar e
   ### Herramientas MCP
 </div>
 
-Tu servidor Search MCP proporciona dos herramientas que los agentes pueden usar:
+Tu servidor Search MCP proporciona tres herramientas que los agentes pueden usar:
 
 - **Search**: Busca en tu sitio para encontrar contenido relevante, devolviendo fragmentos con títulos y enlaces. Úsala para descubrir información o encontrar páginas que coincidan con una consulta.
 - **Query docs filesystem**: Lee y navega el sistema de archivos virtual de tu sitio usando comandos de estilo shell. Úsala para explorar y recuperar contenido, o extraer secciones específicas, incluyendo lecturas de múltiples páginas en una sola llamada.
+- **Submit feedback**: Reporta a tu equipo un problema en la documentación cuando un agente encuentra una página incorrecta, desactualizada, confusa o incompleta. El agente proporciona la ruta de la página y una descripción del problema, que se registra como comentario no útil sobre esa página para que tu equipo pueda corregirla. Esta es la única herramienta que escribe datos; el resto del servidor es de solo lectura.
 
-Los agentes determinan cuándo usar cada herramienta según el contexto de la conversación. Por ejemplo, un agente podría primero buscar en tu sitio para encontrar páginas relevantes y luego usar la herramienta query docs filesystem para leer el contenido completo de los resultados más relevantes.
+Los agentes determinan cuándo usar cada herramienta según el contexto de la conversación. Por ejemplo, un agente podría primero buscar en tu sitio para encontrar páginas relevantes, luego usar la herramienta query docs filesystem para leer el contenido completo de los resultados más relevantes y llamar a submit feedback si detecta un problema con la página durante el proceso.
 
 <div id="mcp-resources">
   ### Recursos MCP

diff --git a/fr/ai/model-context-protocol.mdx b/fr/ai/model-context-protocol.mdx
@@ -37,12 +37,13 @@ Lorsqu'une application d'IA se connecte à votre serveur Search MCP, elle peut r
   ### Outils MCP
 </div>
 
-Votre serveur Search MCP fournit deux outils que les agents peuvent utiliser :
+Votre serveur Search MCP fournit trois outils que les agents peuvent utiliser :
 
 - **Search** : Recherche sur votre site pour trouver du contenu pertinent, en renvoyant des extraits avec des titres et des liens. Utilisez cet outil pour découvrir des informations ou trouver des pages correspondant à une requête.
 - **Query docs filesystem** : Lit et parcourt le système de fichiers virtuel de votre site à l'aide de commandes de type shell. Utilisez cet outil pour parcourir et récupérer du contenu, ou extraire des sections spécifiques—y compris des lectures par lots sur plusieurs pages en un seul appel.
+- **Submit feedback** : Signale un problème de documentation à votre équipe lorsqu'un agent trouve une page incorrecte, obsolète, confuse ou incomplète. L'agent fournit le chemin de la page et une description du problème, qui est enregistrée comme retour négatif sur cette page afin que votre équipe puisse la corriger. C'est le seul outil qui écrit des données ; le reste du serveur est en lecture seule.
 
-Les agents déterminent quand utiliser chaque outil en fonction du contexte de la conversation. Par exemple, un agent peut d'abord rechercher sur votre site pour trouver des pages pertinentes, puis utiliser l'outil query docs filesystem pour lire le contenu complet des résultats les plus pertinents.
+Les agents déterminent quand utiliser chaque outil en fonction du contexte de la conversation. Par exemple, un agent peut d'abord rechercher sur votre site pour trouver des pages pertinentes, puis utiliser l'outil query docs filesystem pour lire le contenu complet des résultats les plus pertinents, et appeler submit feedback s'il découvre un problème avec la page en cours de route.
 
 <div id="mcp-resources">
   ### Ressources MCP

diff --git a/zh/ai/model-context-protocol.mdx b/zh/ai/model-context-protocol.mdx
@@ -36,12 +36,13 @@ Model Context Protocol (MCP，模型上下文协议) 是一个开放协议，用
   ### MCP 工具
 </div>
 
-你的 MCP 服务器提供两个代理可以使用的工具：
+你的 MCP 服务器提供三个代理可以使用的工具：
 
 - **Search**：在你的文档中搜索相关内容，返回带有标题和链接的摘要片段。使用此工具来发现信息或查找与查询匹配的页面。
 - **Query docs filesystem**：使用 shell 风格的命令读取和浏览文档的虚拟文件系统。使用此工具来获取页面内容、浏览文档结构或提取特定部分——包括在单次调用中批量读取多个页面。
+- **Submit feedback**：当代理发现某个页面不正确、过时、令人困惑或不完整时，将文档问题反馈给你的团队。代理会提供页面路径和问题描述，并将其记录为针对该页面的负面反馈，方便你的团队进行修复。这是唯一会写入数据的工具；服务器的其他部分均为只读。
 
-代理会根据对话的上下文来决定何时使用每个工具。例如，代理可能会先搜索你的文档以查找相关页面，然后使用 query docs filesystem 工具读取最相关结果的完整内容。
+代理会根据对话的上下文来决定何时使用每个工具。例如，代理可能会先搜索你的文档以查找相关页面，然后使用 query docs filesystem 工具读取最相关结果的完整内容，并在过程中发现页面存在问题时调用 submit feedback。
 
 <div id="mcp-resources">
   ### MCP 资源