chutes-build 1.2.0 → 1.2.2

package/CHANGELOG.md

@@ -4,6 +4,26 @@ All notable changes to Chutes Build will be documented in this file. The format
 is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/), and this
 project follows [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [1.2.2] - 2026-08-23
+
+### Changed
+
+- **Model routing targets Chutes' native grammar.** The standalone router deployment the virtual `model-router` id pointed at no longer exists; routing is native to the inference host now — the saved-pool alias `default` (with optional `:latency` / `:throughput`), or inline comma-separated pools composed from CHUTES_ROUTING_POOL and CHUTES_ROUTING_STRATEGY. Legacy configs naming `model-router` keep working, and the dead host is removed from every trust list and default.
+
+### Fixed
+
+- **`update --check` works on Windows.** The updater spawned bare `npm`, which does not resolve to `npm.cmd` there, so the version check always reported "program not found".
+- **Models with strict role alternation no longer reject the conversation.** Adjacent same-role system/user messages are merged before sending; Mistral-Nemo-Instruct returned "conversation roles must alternate" before.
+- **`sessions search` queries the local index only.** The CLI shipped every query to a remote registry that answers 404 on Chutes — a wasted credentialed round trip and a warning per search.
+
+## [1.2.1] - 2026-08-22
+
+### Fixed
+
+- **`completions <shell>` now targets the public binary.** Every generated script (bash, zsh, fish, powershell) was built under the upstream binary name, completing calls to a program that does not exist; all four generators derive from one public-name constant and a test pins the absence of the upstream name in the emitted scripts.
+- **`du` works on Windows.** Worktree-home resolution read `$HOME` directly, which is typically unset there, so `chutes-build du` aborted with "neither $CHUTES_BUILD_HOME nor $HOME is set"; the resolver now falls back to `std::env::home_dir()` (`%USERPROFILE%` on Windows), leaving Unix behaviour unchanged.
+- **No upstream placeholders leak into help or completions.** Clap value names derived from internal field ids surfaced as `<XAI_API_BASE_URL>` and `GROK_WS_*` in `--help` and shell completion scripts; those arguments carry explicit value names now, and one MCP doc comment no longer names the upstream binary.
+
 ## [1.2.0] - 2026-08-21
 
 ### Added

package/PRIVACY.md

@@ -56,8 +56,7 @@ explicit hosted feature:
 
 | Destination | Purpose | Data sent |
 | --- | --- | --- |
-| `model-router-ten.vercel.app` | default Chutes-compatible model routing | prompts, selected context/tool schemas, model inputs, Chutes API credential |
-| `llm.chutes.ai` | Chutes inference, model catalog, semantic memory, voice transcription, OCR, and vision | model requests and the selected inputs: prompts/context, memory chunks, audio, images, PDF pages, or sampled video frames; Chutes API credential |
+| `llm.chutes.ai` | Chutes inference (including native model routing via the `model` field), model catalog, semantic memory, voice transcription, OCR, and vision | model requests and the selected inputs: prompts/context, memory chunks, audio, images, PDF pages, or sampled video frames; Chutes API credential |
 | `api.chutes.ai` | Chutes account and media APIs | account requests, media parameters and user-selected media assets; Chutes API credential |
 | `api.chutes.ai/idp` | OAuth sign-in and refresh | authorization code or refresh token, PKCE verifier, registered client ID, and a client secret only when configured |
 | `chutes.ai/docs`, `chutes.ai/news` | mandatory primary-source verification for Chutes topics | requested official page URLs; no Chutes API credential |

package/README.md

@@ -34,7 +34,8 @@ Apache-2.0; the project is not affiliated with or endorsed by SpaceXAI.
 What the fork changes:
 
 - **Chutes, not xAI.** Inference goes to `llm.chutes.ai`, the model catalog is
-  Chutes', `model-router` ("Chutes Auto") is the default route, and media
+  Chutes', the default route is Chutes' native model routing (the saved
+  `default` pool, or an inline pool via `CHUTES_ROUTING_POOL`), and media
   generation, OCR, usage and quota all go through Chutes APIs.
 - **Nothing phones home.** Telemetry, remote error reporting, upload, session
   sharing, remote workspace exposure and self-update are off at compile time and

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "chutes-build",
-  "version": "1.2.0",
+  "version": "1.2.2",
   "description": "Privacy-first, open-source coding agent for the Chutes ecosystem",
   "license": "Apache-2.0",
   "author": "Michael Gasperini (Mikesoft)",
@@ -37,12 +37,12 @@
     "check": "npm test && npm run verify:release && npm pack --dry-run"
   },
   "optionalDependencies": {
-    "chutes-build-darwin-arm64": "1.2.0",
-    "chutes-build-darwin-x64": "1.2.0",
-    "chutes-build-linux-arm64-gnu": "1.2.0",
-    "chutes-build-linux-x64-gnu": "1.2.0",
-    "chutes-build-win32-arm64": "1.2.0",
-    "chutes-build-win32-x64": "1.2.0"
+    "chutes-build-darwin-arm64": "1.2.2",
+    "chutes-build-darwin-x64": "1.2.2",
+    "chutes-build-linux-arm64-gnu": "1.2.2",
+    "chutes-build-linux-x64-gnu": "1.2.2",
+    "chutes-build-win32-arm64": "1.2.2",
+    "chutes-build-win32-x64": "1.2.2"
   },
   "keywords": [
     "ai",