@linxiraos/pi-catalog 1.1.9 → 1.1.11

package/CHANGELOG.md

@@ -2,6 +2,29 @@
 
 ## [Unreleased]
 
+## [1.1.11] - 2026-09-08
+
+### Added
+
+- Added Muse Code as a provider with Muse Spark models and live account-scoped discovery; subscriptions resolve a compact edit-prompt variant, cutting recurring per-request tool bytes without touching other providers. Added Meta's `max` reasoning effort tier to Muse Spark 1.3 (standard) on the Meta Model API and Muse Code.
+
+### Fixed
+
+- GPT-6 Astra keeps its documented 1.05M-token window with `/extended-context` on or off; Codex Astra defaults to 272K, clamps explicit context-window overrides to the server-honored ceiling, uses the documented 922K input cap inside the 1.05M total, and bills long-context requests at the documented 2x input / 1.5x output tier above 272K input (Codex subscription route exempt with free cache writes).
+- Fixed GitHub Copilot enterprise-only model ids inheriting another provider's wire routing (e.g. `gpt-5.6-sol-fast` pinning every request to the `-none` sibling id regardless of thinking level).
+- Extended-context catalog rebuilds resolve each model's maximum window once per process.
+
+## [1.1.10] - 2026-09-07
+
+### Added
+
+- Muse Code provider baseline from the v18.1.12 sync: provider registration, Spark models, live discovery, `max` reasoning tier.
+
+### Fixed
+
+- Fixed OpenCode Go/Zen live model discovery missing `x-opencode-session` attribution: discovery requests carry the stable install id so requests flagged as `Bun fetch` get the required session header.
+- Fixed GPT-6 Astra requests through GitHub Copilot failing with an unsupported endpoint error.
+
 ## [1.1.9] - 2026-09-05
 
 - v18.1.10 sync baseline; retired pi-hashline catalog entry removed from the version-line catalogs.