Merge branch 'main' into mcowger/ollamaContext

Author: chrarnoldus

.changeset/beige-sheep-post.md

@@ -0,0 +1,5 @@
+---
+"kilo-code": minor
+---
+
+Improve Inline Assist model compatibility and performance

.changeset/metal-feet-dream.md

@@ -33,7 +33,7 @@
 			"args": [
 				"--extensionDevelopmentPath=${workspaceFolder}/src",
 				"--disable-extensions",
-				"${workspaceFolder}/launch"
+				"${input:extensionLaunchDir}"
 			],
 			"sourceMaps": true,
 			"outFiles": ["${workspaceFolder}/dist/**/*.js"],
@@ -45,5 +45,13 @@
 			"resolveSourceMapLocations": ["${workspaceFolder}/**", "!**/node_modules/**"],
 			"presentation": { "hidden": false, "group": "tasks", "order": 1 }
 		}
+	],
+	"inputs": [
+		{
+			"id": "extensionLaunchDir",
+			"description": "Directory the dev extension will open in",
+			"default": "${workspaceFolder}/launch",
+			"type": "promptString"
+		}
 	]
 }

.changeset/public-flowers-sell.md

@@ -1,5 +1,31 @@
 # kilo-code
 
+## [v4.85.0]
+
+- [#2119](https://github.com/Kilo-Org/kilocode/pull/2119) [`19dc45d`](https://github.com/Kilo-Org/kilocode/commit/19dc45d1b1578a41c41ecb787e7945513f6554d9) Thanks [@kevinvandijk](https://github.com/kevinvandijk)! - Include changes from Roo Code v3.25.23
+
+    - feat: add custom base URL support for Requesty provider (thanks @requesty-JohnCosta27!)
+    - feat: add DeepSeek V3.1 model to Chutes AI provider (#7294 by @dmarkey, PR by @app/roomote)
+    - Add prompt caching support for Kimi K2 on Groq (thanks @daniel-lxs and @benank!)
+    - Add documentation links for global custom instructions in UI (thanks @app/roomote!)
+    - Ensure subtask results are provided to GPT-5 in OpenAI Responses API
+    - Promote the experimental AssistantMessageParser to the default parser
+    - Update DeepSeek models context window to 128k (thanks @JuanPerezReal)
+    - Enable grounding features for Vertex AI (thanks @anguslees)
+    - Allow orchestrator to pass TODO lists to subtasks
+    - Improved MDM handling
+    - Handle nullish token values in ContextCondenseRow to prevent UI crash (thanks @s97712)
+    - Improved context window error handling for OpenAI and other providers
+    - Add "installed" filter to Marketplace (thanks @semidark)
+    - Improve filesystem access checks (thanks @elianiva)
+    - Add Featherless provider (thanks @DarinVerheijke)
+
+### Patch Changes
+
+- [#2184](https://github.com/Kilo-Org/kilocode/pull/2184) [`0be6743`](https://github.com/Kilo-Org/kilocode/commit/0be6743e08540d1671c10f79b49f17eeac82397e) Thanks [@kevinvandijk](https://github.com/kevinvandijk)! - Fix: add favorite button back to tasks
+
+- [#2125](https://github.com/Kilo-Org/kilocode/pull/2125) [`5828254`](https://github.com/Kilo-Org/kilocode/commit/5828254d47e9073c0f0fc9c9db5ef38eb6358036) Thanks [@nitinprajwal](https://github.com/nitinprajwal)! - Added support to Qwen Code for a custom OAuth credential storage path
+
 ## [v4.84.1]
 
 - [#2113](https://github.com/Kilo-Org/kilocode/pull/2113) [`d40b35a`](https://github.com/Kilo-Org/kilocode/commit/d40b35a3a1efcc2fbfca51d4ca64a8da2aa321e5) Thanks [@chrarnoldus](https://github.com/chrarnoldus)! - The model selector below the chat now shows the correct model list for Qwen Code and some other providers

.vscode/launch.json

@@ -8,6 +8,10 @@ sidebar_label: Free & Budget Models
 
 ## Completely Free Options
 
+### Grok Code Fast 1
+
+This frontier AI model is 100% free in Kilo Code for a limited time. [See the blog post for more details](https://blog.kilocode.ai/p/grok-code-fast-get-this-frontier-ai-model-free).
+
 ### OpenRouter Free Tier Models
 
 OpenRouter offers several models with generous free tiers. **Note:** You'll need to create a free OpenRouter account to access these models.
@@ -20,8 +24,8 @@ OpenRouter offers several models with generous free tiers. **Note:** You'll need
 
 **Available free models:**
 
-- **Qwen3 Coder (free)** - Optimized for agentic coding tasks such as function calling, tool use, and long-context reasoning over repositories. 
-- **Z.AI: GLM 4.5 Air (free)** - Lightweight variant of the GLM-4.5 family, purpose-built for agent-centric applications. 
+- **Qwen3 Coder (free)** - Optimized for agentic coding tasks such as function calling, tool use, and long-context reasoning over repositories.
+- **Z.AI: GLM 4.5 Air (free)** - Lightweight variant of the GLM-4.5 family, purpose-built for agent-centric applications.
 - **DeepSeek: R1 0528 (free)** - Performance on par with OpenAI o1, but open-sourced and with fully open reasoning tokens.
 - **MoonshotAI: Kimi K2 (free)** - Optimized for agentic capabilities, including advanced tool use, reasoning, and code synthesis.
 
@@ -239,8 +243,8 @@ When you need more capability than free models provide, these options deliver ex
 1. **Create OpenRouter account** for free models
 2. **Configure multiple providers** in Kilo Code
 3. **Set up API Configuration Profiles** for easy switching
-5. **Escalate to budget models** when needed
-6. **Reserve premium models** for complex work
+4. **Escalate to budget models** when needed
+5. **Reserve premium models** for complex work
 
 ### Recommended Provider Mix
 

CHANGELOG.md

@@ -6,6 +6,7 @@ export const ghostServiceSettingsSchema = z
 		autoTriggerDelay: z.number().min(1).max(30).default(3).optional(),
 		enableQuickInlineTaskKeybinding: z.boolean().optional(),
 		enableSmartInlineTaskKeybinding: z.boolean().optional(),
+		enableCustomProvider: z.boolean().optional(),
 		apiConfigId: z.string().optional(),
 	})
 	.optional()

apps/kilocode-docs/docs/advanced-usage/free-and-budget-models.md

@@ -3,7 +3,7 @@
 	"displayName": "%extension.displayName%",
 	"description": "%extension.description%",
 	"publisher": "kilocode",
-	"version": "4.84.1",
+	"version": "4.85.0",
 	"icon": "assets/icons/logo-outline-black.png",
 	"galleryBanner": {
 		"color": "#FFFFFF",