fix(executors): synthesize x-opencode-request for custom-named OpenCode providers (#4465) (#4476)

CHANGELOG.md

@@ -30,6 +30,7 @@ _In development — bullets added per PR; finalized at release._
 - **fix(pricing): align Claude Code (`cc`) pricing with current Anthropic per-MTok rates** — the `cc` provider block in the default pricing table had stale numbers across every Claude 4.x family entry — most visibly, `claude-opus-4-5-20251101` was billed at the deprecated Opus 4.1 rate (`input $15` / `output $75`), and `claude-haiku-4-5-20251001` was at half the current Haiku 4.5 rate. The `cached` (cache hit) and `cache_creation` (5-minute cache write) multipliers were also off across Opus 4.6/4.7/4.8, Sonnet 4.5/4.6, Haiku 4.5, and Fable 5. All eight entries now match the rates Anthropic publishes (input, 5m cache write at 1.25x input, cache hit at 0.1x input, output; reasoning billed at the output rate), so cost accounting on the dashboard and per-request usage events stop under- or over-reporting Claude Code spend. (thanks @chulanpro5)
 - **fix(executors): sanitize Anthropic-shape content parts before GitHub Copilot `/chat/completions`** — Claude models on GitHub Copilot driven from clients like Cursor IDE (e.g. `gh/claude-sonnet-4.6`) failed with `Provider returned error: type has to be either 'image_url' or 'text' (reset after 30s)` because the client passed through Anthropic-shape content parts (`tool_use`, `tool_result`, `thinking`) untouched, and the Copilot chat-completions endpoint only accepts `text`/`image_url`. `GithubExecutor.transformRequest` now serializes any unsupported part type as `text` (preserving the model's context), drops empty parts, and collapses to `null` when an assistant message's only content was tool_calls — `tool_calls` ride alongside untouched. Codex-family models still route through `/responses` unchanged. (thanks @cngznNN)
 - **fix(sse):** refactor stall detection to reduce false positives on slow but progressing streams. (thanks @zakirkun)
+- **fix(executors): synthesize `x-opencode-request` for custom-named OpenCode providers** — the OpenCode CLI only emits the `x-opencode-*` header set when the provider id starts with `opencode`; a custom-named provider (e.g. `omniroute`) instead sends `x-session-affinity` / `x-session-id` (mapped to `x-opencode-session` since #4022) but no request-correlation id, so `x-opencode-request` was silently dropped. `OpencodeExecutor` now synthesizes a fresh `x-opencode-request` on that session-affinity fallback path so custom-named providers are not disadvantaged on the opencode.ai upstream. `x-opencode-client` / `x-opencode-project` are intentionally **not** fabricated (no valid client source — an invented value risks upstream rejection) and remain forward-only; `DefaultExecutor` is untouched. ([#4465](https://github.com/diegosouzapw/OmniRoute/issues/4465) — thanks @pizzav-xyz)
 
 ---
 

open-sse/executors/opencode.ts

@@ -1,3 +1,4 @@
+import { randomUUID } from "crypto";
 import {
   BaseExecutor,
   setUserAgentHeader,
@@ -106,6 +107,20 @@ export class OpencodeExecutor extends BaseExecutor {
           findClientHeader("x-session-affinity") || findClientHeader("x-session-id");
         if (sessionAffinity) {
           headers["x-opencode-session"] = sessionAffinity;
+
+          // #4465: a custom-named provider only reaches this fallback because the
+          // OpenCode CLI did NOT emit the x-opencode-* set (it only does so when the
+          // provider id starts with "opencode"). It therefore also dropped
+          // x-opencode-request, a per-request correlation id. Synthesize one so these
+          // users are not disadvantaged versus opencode-prefixed providers on the
+          // opencode.ai upstream. x-opencode-client / x-opencode-project are NOT
+          // fabricated: their valid values are opencode-internal and inventing them
+          // could be rejected upstream — they remain forward-only above. Scoped to this
+          // executor (opencode.ai/zen) and only to the fallback path, so the direct
+          // OpenCode CLI flow (which controls its own request id) is untouched.
+          if (!headers["x-opencode-request"]) {
+            headers["x-opencode-request"] = randomUUID();
+          }
         }
       }
     }

tests/unit/opencode-executor.test.ts

@@ -461,6 +461,71 @@ describe("OpencodeExecutor", () => {
       assert.equal(headers["x-opencode-session"], "sess-go-aff");
     });
   });
+
+  // #4465: custom-named providers reach the session-affinity fallback above, but the
+  // OpenCode CLI never emits x-opencode-request for them (it only emits x-opencode-*
+  // when the provider id starts with "opencode"). Synthesize a request correlation id
+  // so these users are not disadvantaged on the opencode.ai upstream. x-opencode-client
+  // / x-opencode-project are NOT fabricated: their valid values are opencode-internal
+  // and inventing them risks upstream rejection — they stay forward-only.
+  describe("opencode request-id synthesis for custom-named providers (#4465)", () => {
+    it("synthesizes x-opencode-request when only session-affinity is present", () => {
+      const headers = zenExecutor.buildHeaders({ apiKey: "test-key" }, true, {
+        "x-session-affinity": "sess-aff",
+      });
+      assert.equal(headers["x-opencode-session"], "sess-aff");
+      assert.ok(
+        typeof headers["x-opencode-request"] === "string" &&
+          headers["x-opencode-request"].length > 0,
+        "expected a synthesized x-opencode-request id"
+      );
+    });
+
+    it("synthesizes a unique x-opencode-request per call", () => {
+      const a = zenExecutor.buildHeaders({ apiKey: "k" }, true, {
+        "x-session-affinity": "sess-aff",
+      });
+      const b = zenExecutor.buildHeaders({ apiKey: "k" }, true, {
+        "x-session-affinity": "sess-aff",
+      });
+      assert.notEqual(a["x-opencode-request"], b["x-opencode-request"]);
+    });
+
+    it("prefers a client-sent x-opencode-request over the synthesized one", () => {
+      const headers = zenExecutor.buildHeaders({ apiKey: "test-key" }, true, {
+        "x-session-affinity": "sess-aff",
+        "x-opencode-request": "req-real",
+      });
+      assert.equal(headers["x-opencode-request"], "req-real");
+    });
+
+    it("does not fabricate x-opencode-client / x-opencode-project (no client source)", () => {
+      const headers = zenExecutor.buildHeaders({ apiKey: "test-key" }, true, {
+        "x-session-affinity": "sess-aff",
+      });
+      assert.equal(headers["x-opencode-client"], undefined);
+      assert.equal(headers["x-opencode-project"], undefined);
+    });
+
+    it("does not synthesize x-opencode-request on the direct opencode-session path", () => {
+      // opencode CLI (provider id starts with "opencode") sends its own x-opencode-*
+      // set; we must not override/inject when it controls the request id itself.
+      const headers = zenExecutor.buildHeaders({ apiKey: "test-key" }, true, {
+        "x-opencode-session": "direct",
+      });
+      assert.equal(headers["x-opencode-request"], undefined);
+    });
+
+    it("opencode-go executor also synthesizes x-opencode-request on the fallback path", () => {
+      const headers = goExecutor.buildHeaders({ apiKey: "test-key" }, true, {
+        "x-session-affinity": "sess-go-aff",
+      });
+      assert.ok(
+        typeof headers["x-opencode-request"] === "string" &&
+          headers["x-opencode-request"].length > 0
+      );
+    });
+  });
 });
 
 describe("DefaultExecutor", () => {