Fix OpenCode routes: drop LiteLLM-only prompt-cache arg, persist subscription provider for viewer label

2026-08-17 01:29:42 +02:00 · 2026-08-14 17:55:55 +00:00
parent dce70c643a
commit 399c15627b
9 changed files with 60 additions and 27 deletions
@@ -147,3 +147,12 @@ def auth_mode(model_name: str | None) -> str:
    if subscription_model(model_name) or codex.subscription_model(model_name):
        return "subscription"
    return "api_key"
+
+
+def subscription_provider(model_name: str | None) -> str | None:
+    """The subscription behind STRIX_LLM: "opencode", "chatgpt", or None."""
+    if subscription_model(model_name):
+        return PROVIDER
+    if codex.subscription_model(model_name):
+        return "chatgpt"
+    return None
@@ -8,6 +8,7 @@ from typing import TYPE_CHECKING, Any
 from agents.model_settings import ModelSettings
 from openai.types.shared import Reasoning

+from strix.config import opencode
 from strix.config.models import (
    DEFAULT_MODEL_RETRY,
    OPENROUTER_ATTRIBUTION_HEADERS,
@@ -272,6 +273,10 @@ def _prompt_cache_extra_args(model_name: str) -> dict[str, Any] | None:
    """
    if not is_claude_model(model_name):
        return None
+    # OpenCode routes use the raw OpenAI SDK, which rejects this LiteLLM-only
+    # argument; the gateway applies Anthropic prompt caching itself.
+    if opencode.subscription_model(model_name):
+        return None
    if is_bedrock_route(model_name) and not bedrock_route_supports_prompt_caching(model_name):
        return None

@@ -248,6 +248,7 @@ def _persist_run_record(args: argparse.Namespace) -> None:
        "start_time": datetime.now(UTC).isoformat(),
        "end_time": None,
        "auth_mode": opencode.auth_mode(load_settings().llm.model),
+        "subscription_provider": opencode.subscription_provider(load_settings().llm.model),
        "targets_info": args.targets_info,
        "scan_mode": args.scan_mode,
        "instruction": args.instruction,
@@ -101,9 +101,11 @@ export function RunDetails({
  const totalTokens = num(usage.total_tokens);
  const cost = num(usage.cost);
  const subscription = str(raw.auth_mode) === "subscription";
-  const subscriptionLabel = models.some((m) => m.toLowerCase().startsWith("opencode"))
-    ? "OpenCode subscription"
-    : "ChatGPT subscription";
+  const subscriptionProvider =
+    str(raw.subscription_provider) ??
+    (models.some((m) => m.toLowerCase().startsWith("opencode")) ? "opencode" : "chatgpt");
+  const subscriptionLabel =
+    subscriptionProvider === "opencode" ? "OpenCode subscription" : "ChatGPT subscription";

  const sub = (n: number, word: string) => (
    <span className="text-[#666]"> ({formatNumber(n)} {word})</span>
@@ -6,7 +6,7 @@
    <meta name="viewport" content="width=device-width, initial-scale=1.0" />
    <meta name="color-scheme" content="dark" />
    <title>Strix Results</title>
-    <script type="module" crossorigin src="./assets/index-B9ZlcYw2.js"></script>
+    <script type="module" crossorigin src="./assets/index-1LIW3rcB.js"></script>
    <link rel="stylesheet" crossorigin href="./assets/index-DKbLYAbP.css">
  </head>
  <body>
@@ -132,6 +132,7 @@ class ReportState:
            "end_time": None,
            "status": "running",
            "auth_mode": auth_mode,
+            "subscription_provider": opencode.subscription_provider(load_settings().llm.model),
            "targets_info": [],
            "llm_usage": self._build_llm_usage_record(),
        }
@@ -111,6 +111,13 @@ def test_make_model_settings_no_prompt_cache_for_non_claude(model_name: str) ->
    assert make_model_settings(None, model_name=model_name).extra_args is None


+@pytest.mark.parametrize("model_name", ["opencode/claude-sonnet-5", "opencode-go/claude-sonnet-5"])
+def test_no_prompt_cache_for_opencode_claude(model_name: str) -> None:
+    # The OpenCode route uses the raw OpenAI SDK, whose create() rejects the
+    # LiteLLM-only cache_control_injection_points argument.
+    assert _cache_points(model_name) is None
+
+
 def test_no_prompt_cache_for_unmapped_bedrock_claude_model(monkeypatch: Any) -> None:
    # A Bedrock Claude model LiteLLM hasn't mapped must run uncached, not crash.
    unmapped = "bedrock/global.anthropic.claude-brand-new-9"
@@ -94,6 +94,14 @@ def test_auth_mode_covers_both_subscriptions() -> None:
    assert opencode.auth_mode(None) == "api_key"


+def test_subscription_provider() -> None:
+    assert opencode.subscription_provider("opencode/claude-sonnet-5") == "opencode"
+    assert opencode.subscription_provider("opencode-go/kimi-k3") == "opencode"
+    assert opencode.subscription_provider("chatgpt/gpt-5.4") == "chatgpt"
+    assert opencode.subscription_provider("openai/gpt-5.4") is None
+    assert opencode.subscription_provider(None) is None
+
+
 def _response(status_code: int, text: str = "") -> mock.MagicMock:
    response = mock.MagicMock()
    response.status_code = status_code