@inbrace-tech/tokenline 1.2.3 → 1.2.4

package/README.md

@@ -67,7 +67,7 @@ LLMs are stateless — by default, they must reread your entire codebase context
 
 **Prompt Caching** solves this by saving that processed state in the provider's memory:
 - **Cache Write (Cold):** The model reads everything and stores the state. This costs slightly more than base tokens.
-- **Cache Hit (Warm):** If your prompt prefix matches the cached state exactly, the model skips the reading phase. This is **~90% cheaper** (**97.5%** on Claude Fable 5.1 / Mythos 5.1, which bill cache hits at 0.025x base input) and starts responding almost instantly. `tokenline` picks the right multiplier from the active model.
+- **Cache Hit (Warm):** If your prompt prefix matches the cached state exactly, the model skips the reading phase. This is **~90% cheaper** (**97.5%** on Claude Fable 5.1 / Mythos 5.1, which bill cache hits at 0.025x base input, and **95%** on Claude Opus 5.5 at 0.05x) and starts responding almost instantly. `tokenline` picks the right multiplier from the active model.
 - **TTL (Time-To-Live):** The cache operates as a **sliding window**. Every cache hit resets the countdown to its full duration for free. The secret to minimizing rate-limit drain is adjusting your workflow pace to keep the cache continuously **HOT**.
 
 ### 💡 Pro-tip: Forcing a 5m TTL to save rate limits

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@inbrace-tech/tokenline",
-  "version": "1.2.3",
+  "version": "1.2.4",
   "description": "A cache-aware statusline for AI coding CLIs — installs tokenline.sh and wires it into your Claude Code settings.",
   "bin": {
     "tokenline": "dist/cli.js"
@@ -37,6 +37,7 @@
     "access": "public"
   },
   "devDependencies": {
+    "@changesets/changelog-github": "1.0.1",
     "@changesets/cli": "^3.0.2",
     "@eslint/js": "10.0.1",
     "@types/node": "24.13.4",
@@ -63,6 +64,7 @@
     "audit:lockfile": "node --experimental-strip-types scripts/check-lockfile-release-age.mts",
     "test:watch": "vitest",
     "changeset": "changeset",
+    "changeset:check": "node --experimental-strip-types scripts/check-changeset-size.mts",
     "release": "pnpm build && changeset publish"
   }
 }

package/tokenline.sh

@@ -156,13 +156,17 @@ parse_and_prepare_paths() {
     is_gemini=true
   fi
 
-  # Claude Fable 5.1 and Mythos 5.1 bill cache hits at 0.025x base input instead of
-  # the standard 0.1x. Match the API id (claude-fable-5-1[1m]) or the display name
-  # (Fable 5.1) so either field alone is enough.
-  discounted_cache_read=false
-  local discounted_re='(fable|mythos)[- ]5[-.]1([^0-9]|$)'
-  if [[ "${model_id,,}" =~ $discounted_re ]] || [[ "${model,,}" =~ $discounted_re ]]; then
-    discounted_cache_read=true
+  # Cache hits bill at 0.1x base input, except Claude Fable 5.1 / Mythos 5.1 (0.025x)
+  # and Claude Opus 5.5 (0.05x). Match the API id (claude-fable-5-1[1m]) or the
+  # display name (Fable 5.1) so either field alone is enough. Pin the full version:
+  # claude-opus-5 must stay 0.1x, and a later version's price is confirmed, not guessed.
+  claude_read_mult="0.1"
+  local fable_re='(fable|mythos)[- ]5[-.]1([^0-9]|$)'
+  local opus_re='opus[- ]5[-.]5([^0-9]|$)'
+  if [[ "${model_id,,}" =~ $fable_re ]] || [[ "${model,,}" =~ $fable_re ]]; then
+    claude_read_mult="0.025"
+  elif [[ "${model_id,,}" =~ $opus_re ]] || [[ "${model,,}" =~ $opus_re ]]; then
+    claude_read_mult="0.05"
   fi
 
   # Get the current epoch timestamp once to be reused across all calculations
@@ -425,8 +429,7 @@ compute_turn_breakdown() {
       input_mult="1"
       output_mult="4"
     else
-      read_mult="0.1"
-      [ "$discounted_cache_read" = true ] && read_mult="0.025"
+      read_mult="$claude_read_mult"
       write_mult="1.25"
       [ "${ttl_label:-5m}" = "1h" ] && write_mult="2"
       input_mult="1"