move changes to draft spec and use errors to denote rejection of elicitation request

Author: Abhimanyu Siwach

docs/docs.json

@@ -46,7 +46,6 @@
               "docs/concepts/prompts",
               "docs/concepts/tools",
               "docs/concepts/sampling",
-              "docs/concepts/elicitation",
               "docs/concepts/roots",
               "docs/concepts/transports"
             ]
@@ -106,8 +105,7 @@
                 "group": "Client Features",
                 "pages": [
                   "specification/2025-03-26/client/roots",
-                  "specification/2025-03-26/client/sampling",
-                  "specification/2025-03-26/client/elicitation"
+                  "specification/2025-03-26/client/sampling"
                 ]
               },
               {
@@ -204,7 +202,8 @@
                 "group": "Client Features",
                 "pages": [
                   "specification/draft/client/roots",
-                  "specification/draft/client/sampling"
+                  "specification/draft/client/sampling",
+                  "specification/draft/client/elicitation"
                 ]
               },
               {

docs/docs/concepts/elicitation.mdx

@@ -18,17 +18,13 @@ the previous revision, [2024-11-05](/specification/2024-11-05).
 1. Added comprehensive **tool annotations** for better describing tool behavior, like
    whether it is read-only or destructive (PR
    [#185](https://github.com/modelcontextprotocol/specification/pull/185))
-1. Added support for **[elicitation](/specification/2025-03-26/client/elicitation)**, enabling
-   servers to request additional information from users during interactions
 
 ## Other schema changes
 
 - Added `message` field to `ProgressNotification` to provide descriptive status updates
 - Added support for audio data, joining the existing text and image content types
 - Added `completions` capability to explicitly indicate support for argument
   autocompletion suggestions
-- Added `elicitation` capability for clients to indicate support for server-initiated
-  user information requests with array-based responses
 
 See
 [the updated schema](http://github.com/modelcontextprotocol/specification/tree/main/schema/2025-03-26/schema.ts)

docs/specification/2025-03-26/changelog.mdx

@@ -59,10 +59,9 @@ Servers offer any of the following features to clients:
 - **Prompts**: Templated messages and workflows for users
 - **Tools**: Functions for the AI model to execute
 
-Clients may offer the following features to servers:
+Clients may offer the following feature to servers:
 
 - **Sampling**: Server-initiated agentic behaviors and recursive LLM interactions
-- **Elicitation**: Server-initiated requests for additional information from users
 
 ### Additional Utilities
 
@@ -109,14 +108,6 @@ considerations that all implementors must carefully address.
      - What results the server can see
    - The protocol intentionally limits server visibility into prompts
 
-5. **Elicitation Controls**
-   - Users must explicitly approve any elicitation requests
-   - Users should control:
-     - Whether to provide the requested information
-     - What information to share in response
-     - The ability to modify or cancel responses
-   - Clients should clearly indicate which server is requesting information
-
 ### Implementation Guidelines
 
 While MCP itself cannot enforce these security principles at the protocol level,

docs/specification/2025-03-26/index.mdx

@@ -9,12 +9,11 @@ the previous revision, [2025-03-26](/specification/2025-03-26).
 
 1. Removed support for JSON-RPC **[batching](https://www.jsonrpc.org/specification#batch)**
    (PR [#416](https://github.com/modelcontextprotocol/specification/pull/416)) 
-2. TODO
+2. Added support for **[elicitation](/specification/draft/client/elicitation)**, enabling
+   servers to request additional information from users during interactions. (PR [#382](https://github.com/modelcontextprotocol/modelcontextprotocol/pull/382))
 
 ## Other schema changes
 
-- TODO
-
 ## Full changelog
 
 For a complete list of all changes that have been made since the last protocol revision,

docs/specification/draft/changelog.mdx

@@ -2,7 +2,7 @@
 title: Elicitation
 ---
 
-<Info>**Protocol Revision**: 2025-03-26</Info>
+<Info>**Protocol Revision**: draft</Info>
 
 The Model Context Protocol (MCP) provides a standardized way for servers to request additional
 information from users through the client during interactions. This flow allows clients to
@@ -33,7 +33,7 @@ Applications **SHOULD**:
 ## Capabilities
 
 Clients that support elicitation **MUST** declare the `elicitation` capability during
-[initialization](/specification/2025-03-26/basic/lifecycle#initialization):
+[initialization](/specification/draft/basic/lifecycle#initialization):
 
 ```json
 {
@@ -195,17 +195,17 @@ Each content item can be one of:
 
 ## Error Handling
 
-Clients **SHOULD** return errors for common failure cases:
+Clients **SHOULD** return standard JSON-RPC errors for common failure cases:
 
 Example when the user cancels:
 
 ```json
 {
   "jsonrpc": "2.0",
   "id": 1,
-  "result": {
-    "content": [],
-    "cancelled": true
+  "error": {
+    "code": -1,
+    "message": "User rejected the elicitation request"
   }
 }
 ```

docs/specification/2025-03-26/client/elicitation.mdx renamed to docs/specification/draft/client/elicitation.mdx

@@ -59,9 +59,10 @@ Servers offer any of the following features to clients:
 - **Prompts**: Templated messages and workflows for users
 - **Tools**: Functions for the AI model to execute
 
-Clients may offer the following feature to servers:
+Clients may offer the following features to servers:
 
 - **Sampling**: Server-initiated agentic behaviors and recursive LLM interactions
+- **Elicitation**: Server-initiated requests for additional information from users
 
 ### Additional Utilities