@ckyueix/glm-usage 0.1.1 → 0.1.3

package/README.md

@@ -8,6 +8,31 @@ Plan account bound to the session's current provider; clicking expands the
 5-hour / weekly windows, remaining / total quota, reset times and the data
 timestamp in your local timezone.
 
+## 0.1.3
+
+Widens the host requirement from the exact `0.2.1-rc.3` pin to the range
+**`>=0.2.1-rc.2 <0.2.2-0`** — the 0.2.1 series from rc.2 onward, shared by
+`engines.dsh` and the `@deepseek-ai/dsh` peer. Runtime code is unchanged from
+0.1.2; the range reuses the existing rc.2 (0.1.1 production) and rc.3 (0.1.2
+full acceptance) verification, and stays capped below 0.2.2 — hosts at 0.2.2
+or later (including its prereleases) are refused until verified.
+
+## 0.1.2
+
+Updates the host requirement to DeepSeek Harness **0.2.1-rc.3**. The host's
+plugin compatibility gate refuses to load 0.1.1 on rc.3 (its
+`engines.dsh` is pinned to `0.2.1-rc.2`). Runtime code is unchanged from 0.1.1
+— this release only re-pins `engines.dsh`, the `@deepseek-ai/dsh` peer, and the
+`@deepseek-ai/dsh-typert-protocol` runtime dependency to the exact rc.3
+version, and is verified offline against the extracted rc.3 runtime closure
+(real compatibility gate, real Host/Browser Gateway, real typert registry,
+real pi-ai Config, real native settings form; synthetic responses only).
+
+**If you run DSH 0.2.1-rc.2, keep 0.1.1** — 0.1.2 declares exactly
+`0.2.1-rc.3` and deliberately does not widen to untested host versions
+(including future rc.4 or stable 0.2.1). The 0.1.1 parameter-name fix below is
+fully retained.
+
 ## 0.1.1
 
 Preserves the Host method parameter names required by DSH's Remote Gateway.
@@ -26,8 +51,9 @@ package.
 
 ## Requirements
 
-- DeepSeek Harness **0.2.1-rc.2** (`engines.dsh` is enforced; other host
-  versions are untested).
+- DeepSeek Harness **`>=0.2.1-rc.2 <0.2.2-0`** (`engines.dsh`, enforced by the
+  host's compatibility gate; covers 0.2.1-rc.2 / rc.3 / rc.4 / stable 0.2.1 —
+  versions below rc.2 and 0.2.2 or later are untested and refused).
 - A configured GLM Coding Plan provider (Zhipu `open.bigmodel.cn` coding-plan
   routes, e.g. the built-in `zai-coding-cn` China route or explicit
   `openai-completions` profiles).

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@ckyueix/glm-usage",
-  "version": "0.1.1",
+  "version": "0.1.3",
   "description": "GLM Coding Plan usage pill for the DeepSeek Harness composer dock (dual-half: host quota reader + web client entry)",
   "type": "module",
   "license": "SEE LICENSE IN LICENSE",
@@ -17,13 +17,13 @@
     "LICENSE"
   ],
   "engines": {
-    "dsh": "0.2.1-rc.2"
+    "dsh": ">=0.2.1-rc.2 <0.2.2-0"
   },
   "peerDependencies": {
-    "@deepseek-ai/dsh": "0.2.1-rc.2"
+    "@deepseek-ai/dsh": ">=0.2.1-rc.2 <0.2.2-0"
   },
   "dependencies": {
-    "@deepseek-ai/dsh-typert-protocol": "0.2.1-rc.2",
+    "@deepseek-ai/dsh-typert-protocol": "0.2.1-rc.3",
     "@deepseek-ai/schemastery": "~3.18.5"
   },
   "dsh": {