feat: v2.5.12 — plan and quota for every AI tool, Grok Build and Kimi Code, Quit AgentHub (#65)

Usage panel (status bar): - every account shows its plan (Claude subscriptionType, Codex chatgpt_plan_type, Cursor membershipType, Grok subscriptionTier) - new quota rows: Cursor CLI (cursor.com usage-summary), Antigravity (`agy --print /usage`, two model pools; AGENTHUB_AGY_USAGE=0 never runs agy), Grok Build, Kimi Code, OpenCode Go — each login read from the tool's own file, sent only to its own host while valid, never refreshed - in-flight quota reads are shared between tabs AI Tools catalog: Grok Build (`grok`) and Kimi Code (`kimi`) added; Gemini CLI, Hermes and OpenClaw removed. Also: Quit AgentHub (#65, `agenthub stop`), status-area icon, Add to Chat on an editor selection, Settings in three groups, JetBrains-style Changes panel (ticked files/hunks, ignore, blame, file history), terminal font. --no-verify: :3000 is taken by another dev server (lane-detect tests); both suites run by hand — server 2636/2639 (the 3 are :3000), client 1131/1131.
2026-10-08 12:22:14 +07:00
parent 6ef9a5d4b7
commit a9c71cd905
133 changed files with 8007 additions and 1830 deletions
@@ -1,7 +1,8 @@
 /**
 * @file Unit tests for the enhanced pricing calculator and the shared token-usage
 * normalizer: 5m/1h cache-write split, server-tool surcharges, and the per-bucket
- * pricing modifiers (fast mode, US data residency, Batch API).
+ * pricing modifiers (fast mode, US data residency, Batch API), and the
+ * long-context subsets priced at a rule's long_* rates.
 * @author Nguyễn Ngọc Trí Vĩ <vinnt@smartgift.vn>
 */

@@ -363,3 +364,72 @@ describe("DEFAULT_PRICING — Opus 5.5", () => {
    assert.equal(r.total_cost.toFixed(2), "2.72");
  });
 });
+
+describe("long-context tier (prompt > 100k tokens)", () => {
+  // Haiku 5.5: every rate x5 above a 100k-token prompt.
+  const HAIKU = [
+    {
+      model_pattern: "claude-haiku-5-5%",
+      display_name: "Claude Haiku 5.5",
+      input_per_mtok: 0.1,
+      output_per_mtok: 0.5,
+      cache_read_per_mtok: 0.01,
+      cache_write_per_mtok: 0.125,
+      cache_write_1h_per_mtok: 0.2,
+      long_input_per_mtok: 0.5,
+      long_output_per_mtok: 2.5,
+    },
+  ];
+  const haiku = (extra) => bucket({ model: "claude-haiku-5-5", ...extra });
+
+  it("counts a request into the long subsets by its whole prompt: input + cache read + cache write", () => {
+    const short = extractUsageFields({ input_tokens: 100_000, output_tokens: 5 });
+    assert.equal(short.longInput, 0);
+    const long = extractUsageFields({
+      input_tokens: 10,
+      output_tokens: 5,
+      cache_read_input_tokens: 90_000,
+      cache_creation_input_tokens: 10_000,
+      cache_creation: { ephemeral_5m_input_tokens: 4_000, ephemeral_1h_input_tokens: 6_000 },
+    });
+    assert.deepEqual(
+      [
+        long.longInput,
+        long.longOutput,
+        long.longCacheRead,
+        long.longCacheWrite,
+        long.longCacheWrite1h,
+      ],
+      [10, 5, 90_000, 10_000, 6_000]
+    );
+  });
+
+  it("prices the long subset at the long rates, cache rates scaled by the input ratio", () => {
+    const all = {
+      input_tokens: M,
+      output_tokens: M,
+      cache_read_tokens: M,
+      cache_write_tokens: 2 * M,
+      cache_write_1h_tokens: M,
+    };
+    const long = Object.fromEntries(Object.entries(all).map(([k, v]) => [`long_${k}`, v]));
+    // 0.5 + 2.5 + 0.05 + 0.625 + 1.0
+    assert.equal(calculateCost([haiku({ ...all, ...long })], HAIKU).total_cost, 4.675);
+  });
+
+  it("prices only the long part of a bucket at the long rates", () => {
+    const r = calculateCost([haiku({ input_tokens: 2 * M, long_input_tokens: M })], HAIKU);
+    assert.equal(r.total_cost, 0.6);
+    assert.equal(r.breakdown[0].long_input_tokens, M);
+  });
+
+  it("never prices more long tokens than the bucket holds", () => {
+    const r = calculateCost([haiku({ input_tokens: M, long_input_tokens: 3 * M })], HAIKU);
+    assert.equal(r.total_cost, 0.5);
+  });
+
+  it("prices a long subset at the standard rates when the rule has no long rates", () => {
+    const r = calculateCost([bucket({ input_tokens: M, long_input_tokens: M })], RULES);
+    assert.equal(r.total_cost, 5);
+  });
+});