Merge pull request modelcontextprotocol#93 from evalstate/feature/audio-modality

Updated Specification and Documentation to support Audio Modality.

Author: dsp-ant

docs/specification/draft/client/sampling.md

@@ -8,7 +8,7 @@ weight: 40
 **Protocol Revision**: draft
 {{< /callout >}}
 
-The Model Context Protocol (MCP) provides a standardized way for servers to request LLM sampling ("completions" or "generations") from language models via clients. This flow allows clients to maintain control over model access, selection, and permissions while enabling servers to leverage AI capabilities&mdash;with no server API keys necessary. Servers can request text or image-based interactions and optionally include context from MCP servers in their prompts.
+The Model Context Protocol (MCP) provides a standardized way for servers to request LLM sampling ("completions" or "generations") from language models via clients. This flow allows clients to maintain control over model access, selection, and permissions while enabling servers to leverage AI capabilities&mdash;with no server API keys necessary. Servers can request text, audio, or image-based interactions and optionally include context from MCP servers in their prompts.
 
 ## User Interaction Model
 
@@ -27,7 +27,7 @@ Implementations are free to expose sampling through any interface pattern that s
 
 ## Capabilities
 
-Clients that support sampling **MUST** declare the `sampling` capability during [initialization]({{< ref "/specification/draft/basic/lifecycle#initialization" >}}):
+Clients that support sampling **MUST** declare the `sampling` capability during [initialization]({{< ref "../basic/lifecycle#initialization" >}}):
 
 ```json
 {
@@ -142,6 +142,16 @@ Sampling messages can contain:
 }
 ```
 
+#### Audio Content
+```json
+{
+  "type": "audio",
+  "data": "base64-encoded-audio-data",
+  "mimeType": "audio/wav"
+}
+```
+
+
 ### Model Preferences
 
 Model selection in MCP requires careful abstraction since servers and clients may use different AI providers with distinct model offerings. A server cannot simply request a specific model by name since the client may not have access to that exact model or may prefer to use a different provider's equivalent model.

docs/specification/draft/server/prompts.md

@@ -189,6 +189,17 @@ Image content allows including visual information in messages:
 ```
 The image data **MUST** be base64-encoded and include a valid MIME type. This enables multi-modal interactions where visual context is important.
 
+#### Audio Content
+Audio content allows including audio information in messages:
+```json
+{
+  "type": "audio",
+  "data": "base64-encoded-audio-data",
+  "mimeType": "audio/wav"
+}
+```
+The audio data MUST be base64-encoded and include a valid MIME type. This enables multi-modal interactions where audio context is important.
+
 #### Embedded Resources
 Embedded resources allow referencing server-side resources directly in messages:
 ```json

docs/specification/draft/server/tools.md

@@ -188,6 +188,15 @@ Tool results can contain multiple content items of different types:
 }
 ```
 
+#### Audio Content
+```json
+{
+  "type": "audio",
+  "data": "base64-encoded-audio-data",
+  "mimeType": "audio/wav"
+}
+```
+
 #### Embedded Resources
 
 [Resources]({{< ref "/specification/draft/server/resources" >}}) **MAY** be embedded, to provide additional context or data, behind a URI that can be subscribed to or fetched again by the client later:

package.json

@@ -11,8 +11,8 @@
     "node": ">=20"
   },
   "scripts": {
-    "validate:schema": "tsc --noEmit schema/schema.ts",
-    "generate:json": "typescript-json-schema --defaultNumberType integer --required schema/schema.ts \"*\" -o schema/schema.json",
+    "validate:schema": "tsc --noEmit schema/schema.ts schema/draft/schema.ts",
+    "generate:json": "typescript-json-schema --defaultNumberType integer --required schema/schema.ts \"*\" -o schema/schema.json && typescript-json-schema --defaultNumberType integer --required schema/draft/schema.ts \"*\" -o schema/draft/schema.json",
     "serve:docs": "hugo --source site/ server --logLevel debug --disableFastRender"
   },
   "devDependencies": {