chutes-build 2.0.5 → 2.0.7

package/CHANGELOG.md

@@ -4,6 +4,67 @@ 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).
 
+## [2.0.7] - 2026-10-27
+
+### Changed
+
+- Auto now pools tool-capable PAYG models together with TEE models. The
+  subscription-only split is gone: a catalogue row no longer has to be
+  confidential to join Auto. PAYG rows keep their `PAYG` notice, and a dry
+  balance still surfaces Chutes' 402 instead of being hidden.
+- Tool support for the PAYG beta rows is curated from endpoint probes
+  (`DeepSeek-V4-Pro`, `DeepSeek-V4.1-Flash`, `Qwen3.8-Max`, `Qwen3.8-2.4T-A95B`,
+  `GLM-5.3`, `GLM-5.2-Fast`, `Kimi-K2.7-Code`), because the catalogue still
+  omits `supported_features` for them.
+
+## [2.0.6] - 2026-10-26
+
+### Added
+
+- `chutes-build routing` and the `/routing` slash command show the saved
+  Chutes model-routing pool, the three strategies, and live TTFT/TPS from
+  `GET /invocations/stats/llm`. Pool members and stats are resolved by
+  `chute_id` against the live catalogue, so the view shows model names rather
+  than private chute labels.
+- The routing view is read-only. Alias writes are not a supported surface yet:
+  the Chutes endpoint is eventually consistent and intermittently answers 502,
+  so the write path stays hidden until the platform stabilises.
+- When the account has a saved `default` routing pool, Auto targets
+  `default:latency` instead of composing an inline pool, so the server-side
+  routing choice made in the Chutes web app is honoured.
+- Reasoning-effort controls are exposed only for models whose accepted values
+  were probed against the Chutes endpoint. The catalogue publishes the
+  `reasoning` feature but not the effort menu, so the verified value set is used
+  instead of the generic fallback.
+- The external-provider `deepseek-ai/DeepSeek-V4-Pro-0813` and
+  `Qwen/Qwen3.8-Max-0902` rows carry no feature list, so their effort menus come
+  from endpoint-verified values (`low|medium|high` and `low|medium|xhigh`).
+  `DeepSeek-V4-Flash` and `Kimi-K3` use their verified value sets too. The
+  2026-10-26 beta wave adds `DeepSeek-V4.1-Flash` (`low|medium|high`), `GLM-5.3`
+  and `GLM-5.2-Fast` (`low|high|max`), `Kimi-K2.7-Code`
+  (`low|medium|high|max`), and `Qwen3.8-2.4T-A95B` (`low|medium|xhigh`). All of
+  these are PAYG-only non-TEE rows; Auto still leaves them out, so selecting
+  them is explicit and never spends PAYG silently.
+
+### Fixed
+
+- The live invocation-stats body is tens of megabytes: fetch it with a longer
+  deadline and `Accept-Encoding: identity` so the routing view can decode it
+  behind edges that negotiate an unsupported compression.
+- `/routing` does not block the TUI on the stats fetch: a cold or stale cache
+  starts a background refresh and asks the user to run the command again.
+- Auto excludes `confidential_compute: false` models explicitly instead of
+  relying on their missing feature list, so a PAYG model can never join the
+  automatic pool.
+- The saved-pool alias cache is seeded on every interactive startup path, not
+  only the headless one, so Auto can honour `default:latency` in the TUI and
+  leader modes.
+- Credential headers are marked sensitive, so keys cannot appear in HTTP debug
+  logs even for management routes that need the raw-key form.
+- The zstd stack is on the patched 0.14 line (RUSTSEC-2026-0391, -0398,
+  -0403, -0410) and the unused tonic zstd feature is gone; `cargo deny` is
+  clean again.
+
 ## [2.0.5] - 2026-10-25
 
 ### Fixed

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "chutes-build",
-  "version": "2.0.5",
+  "version": "2.0.7",
   "description": "Privacy-first, open-source coding agent for the Chutes ecosystem",
   "license": "Apache-2.0",
   "author": { "name": "Michael Gasperini", "email": "michael@mikesoft.it", "url": "https://mikesoft.it" },
@@ -37,12 +37,12 @@
     "check": "npm test && npm run verify:release && npm pack --dry-run"
   },
   "optionalDependencies": {
-    "chutes-build-darwin-arm64": "2.0.5",
-    "chutes-build-darwin-x64": "2.0.5",
-    "chutes-build-linux-arm64-gnu": "2.0.5",
-    "chutes-build-linux-x64-gnu": "2.0.5",
-    "chutes-build-win32-arm64": "2.0.5",
-    "chutes-build-win32-x64": "2.0.5"
+    "chutes-build-darwin-arm64": "2.0.7",
+    "chutes-build-darwin-x64": "2.0.7",
+    "chutes-build-linux-arm64-gnu": "2.0.7",
+    "chutes-build-linux-x64-gnu": "2.0.7",
+    "chutes-build-win32-arm64": "2.0.7",
+    "chutes-build-win32-x64": "2.0.7"
   },
   "keywords": [
     "ai",