Assistant: Use user selected model (#9227) (#9231)

- Cherry-picks #9227 to main
- Addresses #9226

Description from @georgestagg's original PR below 👇 

---

Addresses #9226.

I think this broke when the Code OSS language model proposed API changed
to introduce `prepareLanguageModelChat()`.

These changes essentially defer selection of the model ID until the very
last second in e.g. `provideLanguageModelChatResponse()` and other
methods, rather than trying to store a specific model ID on a class
property such as `this._config` or `this.model`, since a single language
model provider instance can now handle multiple model IDs.

## QA

* Open Positron
* Open Assistant, select a non-default model for your provider, e.g.
Claude 4 Opus.
* Send a request, watching the Assistant log output pane.
* Ensure that the right model is selected by watching the model ID
reported in the logs.

* Repeat for the Anthropic and the AWS Bedrock providers, to cover both
the Anthropic SDK and Vercel AI SDK code paths.

* We should probably also check the `maxOutputTokens` config setting
still works, since I've had to tweak that code.

<img width="950" height="63" alt="Screenshot 2025-08-29 at 14 50 24"
src="https://github.com/user-attachments/assets/82585446-7dac-46cd-9255-fc187647ddb4"
/>
<img width="945" height="59" alt="Screenshot 2025-08-29 at 14 50 51"
src="https://github.com/user-attachments/assets/364d1270-9f03-44af-84de-d19d7ed63761"
/>

Co-authored-by: George Stagg <[email protected]>

Author: sharon-wang

build/secrets/.secrets.baseline

@@ -90,6 +90,10 @@
     {
       "path": "detect_secrets.filters.allowlist.is_line_allowlisted"
     },
+    {
+      "path": "detect_secrets.filters.common.is_baseline_file",
+      "filename": "build/secrets/.secrets.baseline"
+    },
     {
       "path": "detect_secrets.filters.heuristic.is_indirect_reference"
     },
@@ -1051,7 +1055,7 @@
         "filename": "extensions/positron-assistant/src/models.ts",
         "hashed_secret": "ba10d824a2c4c13ce293171d14843c0f0043d039",
         "is_verified": false,
-        "line_number": 819,
+        "line_number": 833,
         "is_secret": false
       }
     ],
@@ -1734,5 +1738,5 @@
       }
     ]
   },
-  "generated_at": "2025-08-25T17:39:02Z"
+  "generated_at": "2025-08-29T13:42:08Z"
 }

extensions/positron-assistant/package-lock.json

@@ -575,7 +575,7 @@
     "@types/mocha": "^9.1.0",
     "@types/node": "^20",
     "@types/sinon": "^17.0.3",
-    "ai": "^4.1.46",
+    "ai": "^4.3.19",
     "eslint": "^9.13.0",
     "google-auth-library": "^9.15.1",
     "mocha": "^9.2.1",

extensions/positron-assistant/package.json

@@ -128,7 +128,7 @@ export class AnthropicLanguageModel implements positron.ai.LanguageModelChatProv
 		const anthropicMessages = toAnthropicMessages(messages);
 
 		const body: Anthropic.MessageStreamParams = {
-			model: this._config.model,
+			model: model.id,
 			max_tokens: options.modelOptions?.maxTokens ?? this.maxOutputTokens,
 			tools,
 			tool_choice,
@@ -140,7 +140,7 @@ export class AnthropicLanguageModel implements positron.ai.LanguageModelChatProv
 
 		// Log request information - the request ID is only available upon connection.
 		stream.on('connect', () => {
-			log.info(`[anthropic] Start request ${stream.request_id} to ${this._config.model}: ${anthropicMessages.length} messages`);
+			log.info(`[anthropic] Start request ${stream.request_id} to ${model.id}: ${anthropicMessages.length} messages`);
 			if (log.logLevel <= vscode.LogLevel.Trace) {
 				log.trace(`[anthropic] SEND messages.stream [${stream.request_id}]: ${JSON.stringify(body, null, 2)}`);
 			} else {
@@ -280,7 +280,7 @@ export class AnthropicLanguageModel implements positron.ai.LanguageModelChatProv
 			}
 		}
 		const result = await this._client.messages.countTokens({
-			model: this._config.model,
+			model: model.id,
 			messages,
 		});
 		return result.input_tokens;