Never stop coding. Free MIT AI gateway: one endpoint, 375 providers (150+ free), 1,400+ models Kimi, Claude, GPT, Gemini, GLM, DeepSeek, MiniMax. Works with Claude Code, Codex, Cursor, OpenCode, Cline & Copilot. Quota-aware auto-fallback, RTK+Caveman compression saves 15-95% tokens, MCP/A2A, Desktop/PWA. Built by 900+ contributors
TypeScript a2aai-agentsai-gatewayanthropicclaudeclaude-codeclinecodexcopilotcursordeepseekfree-aigeminikimillm-gatewaymcpopenaiopenai-proxyqwentoken-saver
chore(release): finalize v3.8.11 changelog + repair release-gate test drift
diegosouzapw75bccccbef64
diff --git a/.env.example b/.env.exampleindex 252ee8d853..d482763231 100644--- a/.env.example+++ b/.env.example@@ -585,6 +585,11 @@ CODEX_OAUTH_CLIENT_ID=app_EMoamEEZ73f0CkXaXp7hrann# Trae OAuth token override. Used by: open-sse/executors/trae.ts.# TRAE_TOKEN=+# ── The Old LLM (theoldllm) ──+# Playwright navigation timeout (ms) for the browser-backed token capture.+# Used by: open-sse/executors/theoldllm.ts. Default: 30000 (30s).+# THEOLDLLM_NAV_TIMEOUT_MS=30000+# ── Gemini / Gemini CLI / Antigravity / Windsurf (all Google-based) ──# These providers ship public OAuth client_id/secret values (or Firebase Web# keys) embedded in their public CLIs/binaries. Defaults are baked into thediff --git a/CHANGELOG.md b/CHANGELOG.mdindex e04cd1f272..76dc95a9b5 100644--- a/CHANGELOG.md+++ b/CHANGELOG.md@@ -4,9 +4,7 @@----## [3.8.11] — Unreleased--_Development cycle in progress — entries are added as work merges into `release/v3.8.11` and finalized by the release flow._+## [3.8.11] — 2026-06-05### ✨ New Features@@ -44,6 +42,7 @@ _Development cycle in progress — entries are added as work merges into `releas- **build:** finish the build-output-isolation cleanup — `assembleStandalone.mjs` now derives both its async (`syncStandalone*`) and sync copy paths from a single `NATIVE_ASSET_ENTRIES`/`EXTRA_MODULE_ENTRIES` source of truth (previously two hand-maintained lists that could silently drift), guarded by a new parity test; and the `Dockerfile` drops 5 redundant per-module `COPY` overrides (`@swc/helpers`, `pino-abstract-transport`, `pino-pretty`, `split2`, `migrations`) now that `assembleStandalone` bundles them into the standalone regardless of NFT/Turbopack tracing (validated with a real Turbopack `docker build` + boot → `/api/monitoring/health` 200; `better-sqlite3` stays explicit since only its native `build/` is synced) (#3187 — thanks @diegosouzapw)- **combo:** add a regression guard asserting the same-provider cascade is short-circuited by the connection-cooldown layer (#3200 — thanks @diegosouzapw)+- **repo:** housekeeping — ignore the generated `coverage/` output dir and prune deprecated `.agents/skills/*` SKILL definitions superseded by the current workflow skills (thanks @diegosouzapw)### 🙌 Contributorsdiff --git a/docs/reference/ENVIRONMENT.md b/docs/reference/ENVIRONMENT.mdindex 67e2b6156a..5b7ecced97 100644--- a/docs/reference/ENVIRONMENT.md+++ b/docs/reference/ENVIRONMENT.md@@ -249,6 +249,7 @@ OmniRoute provides a two-layer defense: request-side injection scanning and resp| `OMNIROUTE_PUBLIC_BASE_URL` | _(unset)_ | `open-sse/executors/chatgpt-web.ts` | Browser-facing OmniRoute origin used for image URLs in API responses (e.g., `/v1/chatgpt-web/image/<id>`). Set this when OpenWebUI or another relay reaches OmniRoute by an internal URL but the user's browser must fetch images from a LAN, tunnel, or public origin. Do **not** include `/v1`. || `OMNIROUTE_CGPT_WEB_IMAGE_TIMEOUT_MS` | `180000` (3 min) | `open-sse/executors/chatgpt-web.ts` | Max wait time for an async chatgpt-web image to land via the celsius WebSocket. Increase during upstream queue-deep windows. || `OMNIROUTE_CGPT_WEB_IMAGE_CACHE_MAX_MB` | `256` | `open-sse/services/chatgptImageCache.ts` | Total in-memory byte budget (MB) for the chatgpt-web image cache serving `/v1/chatgpt-web/image/<id>`. Lower on memory-constrained hosts; raise if image generation is heavy and clients race the 30-minute TTL. |+| `THEOLDLLM_NAV_TIMEOUT_MS` | `30000` (30s) | `open-sse/executors/theoldllm.ts` | Playwright navigation timeout (ms) for the browser-backed token capture used by the The Old LLM (theoldllm) free provider. Raise on slow networks if the relay page is slow to settle. || `KIE_CALLBACK_URL` | _(unset)_ | `open-sse/utils/kieTask.ts` | Public callback URL for asynchronous kie.ai jobs. Highest-priority override before `OMNIROUTE_KIE_CALLBACK_URL` and `OMNIROUTE_PUBLIC_URL`. || `OMNIROUTE_KIE_CALLBACK_URL` | _(unset)_ | `open-sse/utils/kieTask.ts` | Alternate spelling of `KIE_CALLBACK_URL`. Falls back when the primary variable is unset. || `OMNIROUTE_PUBLIC_URL` | _(unset)_ | `open-sse/utils/kieTask.ts` | Public origin used to compose async callback URLs. Lowest-priority fallback for kie.ai callbacks; also used as a generic public URL for other relays. |diff --git a/tests/unit/i18n-nest-dotted-keys.test.ts b/tests/unit/i18n-nest-dotted-keys.test.tsindex 40878e2680..c4d1ab33a8 100644--- a/tests/unit/i18n-nest-dotted-keys.test.ts+++ b/tests/unit/i18n-nest-dotted-keys.test.ts@@ -1,7 +1,9 @@import test from "node:test";import assert from "node:assert/strict";-import { nestDottedKeys } from "@/i18n/request";+// The shipped helper is `normalizeComplianceEventTypes` (#3185); it nests dotted+// keys under `compliance.eventTypes` and is a no-op for messages without that path.+import { normalizeComplianceEventTypes as nestDottedKeys } from "@/i18n/request";test("nestDottedKeys expands flat compliance.eventTypes keys into nested objects", () => {const input = {diff --git a/tests/unit/model-capabilities-registry.test.ts b/tests/unit/model-capabilities-registry.test.tsindex bf0adc4e63..559df1eafd 100644--- a/tests/unit/model-capabilities-registry.test.ts+++ b/tests/unit/model-capabilities-registry.test.ts@@ -68,9 +68,10 @@ test("canonical model capability resolver lets exact synced metadata override gl}),},antigravity: {- // The resolver returns "gemini-3.1-pro-high" unchanged (ANTIGRAVITY_MODEL_ALIASES only maps- // the public-facing alias → internal, not the reverse). Save under the canonical resolved key.- "gemini-3.1-pro-high": buildCapability({+ // Since #3229, ANTIGRAVITY_MODEL_ALIASES maps both "gemini-3.1-pro-high" and+ // "gemini-3.1-pro-low" → "gemini-3.1-pro", so the capability resolver looks+ // synced metadata up under the canonical "gemini-3.1-pro" key. Save it there.+ "gemini-3.1-pro": buildCapability({tool_call: false,reasoning: false,modalities_input: JSON.stringify(["text"]),diff --git a/tests/unit/nvidia-nim-validator.test.ts b/tests/unit/nvidia-nim-validator.test.tsindex 1150e26aa7..d20eb112dc 100644--- a/tests/unit/nvidia-nim-validator.test.ts+++ b/tests/unit/nvidia-nim-validator.test.ts@@ -1,99 +1,107 @@import test from "node:test";import assert from "node:assert/strict";+import http from "node:http";// #2463 — NVIDIA NIM validation must not crash with `e.startsWith is not a function`// when providerSpecificData has malformed shapes; and the validation must use a// direct chat probe instead of the /models probe.+//+// #3226 — the validator now probes via `directHttpsRequest` → `safeOutboundFetch`+// with `bypassProxyPatch: true`, which uses the ORIGINAL (un-patched) native fetch+// captured at module load. Patching `globalThis.fetch` in the test no longer+// intercepts it, so we point `baseUrl` at a real local HTTP server instead (the+// outbound guard is "none" for validation, so 127.0.0.1 is reachable). This+// exercises the true code path end-to-end without a real upstream call.+//+// IMPORTANT: import the validator at FILE LOAD (top-level), not lazily inside a+// test. `proxyFetch` captures the un-patched `globalThis.fetch` on its first+// import; if the first import happened inside a test that had already patched+// `globalThis.fetch`, `getOriginalFetch()` would forever return that test's mock+// and poison every later bypass call. Loading here pins the real native fetch.+const { validateProviderApiKey } = await import("../../src/lib/providers/validation.ts");-test("normalizeBaseUrl tolerates non-string baseUrl without throwing", async () => {- // Indirect probe — call validation entrypoint with a non-string baseUrl in PSD;- // the function should return a normal Validation result (not throw a TypeError- // such as `e.startsWith is not a function` after minification — see #2463).- const origFetch = globalThis.fetch;- globalThis.fetch = (async () =>- new Response(JSON.stringify({}), {- status: 400,- headers: { "content-type": "application/json" },- })) as typeof fetch;-+async function withMockServer(+ handler: (req: http.IncomingMessage, res: http.ServerResponse) => void,+ fn: (baseUrl: string) => Promise<void>+): Promise<void> {+ const server = http.createServer(handler);+ await new Promise<void>((resolve) => server.listen(0, "127.0.0.1", resolve));+ const address = server.address();+ const port = typeof address === "object" && address ? address.port : 0;+ const baseUrl = `http://127.0.0.1:${port}/v1`;try {- const { validateProviderApiKey } = await import("../../src/lib/providers/validation.ts");- const result = await validateProviderApiKey({- provider: "nvidia",- apiKey: "nv-test-key",- // Malformed: baseUrl is an object instead of a string. Pre-fix this would- // crash inside normalizeBaseUrl with `.trim is not a function`.- providerSpecificData: { baseUrl: { not: "a string" } as any },- });- assert.equal(typeof result, "object");- assert.equal(typeof result.valid, "boolean");- // Whether the call succeeds (with default baseUrl) or fails gracefully with- // an outbound URL guard error is fine — the contract is "no TypeError leak".- if (!result.valid && typeof result.error === "string") {- assert.ok(- !result.error.includes("startsWith"),- `error must not mention startsWith TypeError, got: ${result.error}`- );- assert.ok(- !result.error.includes("is not a function"),- `error must not mention TypeError, got: ${result.error}`- );- }+ await fn(baseUrl);} finally {- globalThis.fetch = origFetch;+ await new Promise<void>((resolve) => server.close(() => resolve()));+ }+}++test("normalizeBaseUrl tolerates non-string baseUrl without throwing", async () => {+ // Call validation entrypoint with a non-string baseUrl in PSD; the function+ // should return a normal Validation result (not throw a TypeError such as+ // `e.startsWith is not a function` after minification — see #2463). A malformed+ // baseUrl normalizes to "" and yields an invalid relative probe URL, which the+ // outbound layer rejects gracefully — so no real upstream call is made.+ const result = await validateProviderApiKey({+ provider: "nvidia",+ apiKey: "nv-test-key",+ providerSpecificData: { baseUrl: { not: "a string" } as any },+ });+ assert.equal(typeof result, "object");+ assert.equal(typeof result.valid, "boolean");+ if (!result.valid && typeof result.error === "string") {+ assert.ok(+ !result.error.includes("startsWith"),+ `error must not mention startsWith TypeError, got: ${result.error}`+ );+ assert.ok(+ !result.error.includes("is not a function"),+ `error must not mention TypeError, got: ${result.error}`+ );}});test("nvidia specialty validator returns Invalid API key on 401", async () => {- const origFetch = globalThis.fetch;- globalThis.fetch = (async () =>- new Response(JSON.stringify({ error: "unauthorized" }), {- status: 401,- headers: { "content-type": "application/json" },- })) as typeof fetch;-- try {- const { validateProviderApiKey } = await import("../../src/lib/providers/validation.ts");- const result = await validateProviderApiKey({- provider: "nvidia",- apiKey: "nv-badkey",- providerSpecificData: {},- });- assert.equal(result.valid, false);- assert.equal(result.error, "Invalid API key");- } finally {- globalThis.fetch = origFetch;- }+ await withMockServer(+ (_req, res) => {+ res.writeHead(401, { "content-type": "application/json" });+ res.end(JSON.stringify({ error: "unauthorized" }));+ },+ async (baseUrl) => {+ const result = await validateProviderApiKey({+ provider: "nvidia",+ apiKey: "nv-badkey",+ providerSpecificData: { baseUrl },+ });+ assert.equal(result.valid, false);+ assert.equal(result.error, "Invalid API key");+ }+ );});test("nvidia specialty validator skips /models probe entirely", async () => {const calls: string[] = [];- const origFetch = globalThis.fetch;- globalThis.fetch = (async (url: any) => {- calls.push(String(url));- return new Response(JSON.stringify({}), {- status: 200,- headers: { "content-type": "application/json" },- });- }) as typeof fetch;-- try {- const { validateProviderApiKey } = await import("../../src/lib/providers/validation.ts");- const result = await validateProviderApiKey({- provider: "nvidia",- apiKey: "nv-key",- providerSpecificData: {},- });- assert.equal(result.valid, true);- assert.ok(- calls.every((u) => !u.endsWith("/v1/models")),- `should not call /v1/models, called: ${JSON.stringify(calls)}`- );- assert.ok(- calls.some((u) => u.endsWith("/chat/completions")),- `should call /chat/completions, called: ${JSON.stringify(calls)}`- );- } finally {- globalThis.fetch = origFetch;- }+ await withMockServer(+ (req, res) => {+ calls.push(String(req.url));+ res.writeHead(200, { "content-type": "application/json" });+ res.end(JSON.stringify({}));+ },+ async (baseUrl) => {+ const result = await validateProviderApiKey({+ provider: "nvidia",+ apiKey: "nv-key",+ providerSpecificData: { baseUrl },+ });+ assert.equal(result.valid, true);+ assert.ok(+ calls.every((u) => !u.endsWith("/v1/models")),+ `should not call /v1/models, called: ${JSON.stringify(calls)}`+ );+ assert.ok(+ calls.some((u) => u.endsWith("/chat/completions")),+ `should call /chat/completions, called: ${JSON.stringify(calls)}`+ );+ }+ );});diff --git a/tests/unit/synced-model-delete-persist-3199.test.ts b/tests/unit/synced-model-delete-persist-3199.test.tsindex db32ee1f42..7e9271da2d 100644--- a/tests/unit/synced-model-delete-persist-3199.test.ts+++ b/tests/unit/synced-model-delete-persist-3199.test.ts@@ -9,13 +9,22 @@*/import test, { before, after } from "node:test";import assert from "node:assert/strict";+import os from "node:os";+import path from "node:path";+import fs from "node:fs";-import {- replaceSyncedAvailableModelsForConnection,- getSyncedAvailableModels,- mergeModelCompatOverride,-} from "../../src/lib/localDb.ts";-import { resetDbInstance } from "../../src/lib/db/core.ts";+// Hermetic DB: this test marks a model hidden (an override persisted in the+// `modelCompatOverrides` key_value namespace). Without an isolated DATA_DIR it+// would write that override into the shared dev/CI database and never clean it+// up, so the SECOND run would see `model-del` already hidden and the first-sync+// precondition would fail. Point DATA_DIR at a throwaway dir before any import+// that opens the SQLite connection.+const tmpDir = fs.mkdtempSync(path.join(os.tmpdir(), "omniroute-test-synced-del-"));+process.env.DATA_DIR = tmpDir;++const { replaceSyncedAvailableModelsForConnection, getSyncedAvailableModels, mergeModelCompatOverride } =+ await import("../../src/lib/localDb.ts");+const { resetDbInstance } = await import("../../src/lib/db/core.ts");before(() => {resetDbInstance();@@ -23,6 +32,7 @@ before(() => {after(() => {resetDbInstance();+ fs.rmSync(tmpDir, { recursive: true, force: true });});test("a hidden (deleted) synced model is not re-added on re-import", async () => {diff --git a/tests/unit/web-session-credentials.test.ts b/tests/unit/web-session-credentials.test.tsindex 8f715d0088..9f83f8910f 100644--- a/tests/unit/web-session-credentials.test.ts+++ b/tests/unit/web-session-credentials.test.ts@@ -15,10 +15,11 @@ test("web session credential metadata covers every web-cookie provider", () => {});test("web session credential metadata identifies cookie, token, and no-auth providers", () => {+ // Grok needs BOTH sso and sso-rw cookies (#3180)assert.deepEqual(webSessionCredentials.getWebSessionCredentialRequirement("grok-web"), {kind: "cookie",- credentialName: "sso",- placeholder: "sso=...",+ credentialName: "sso + sso-rw",+ placeholder: "sso=...; sso-rw=...",acceptsFullCookieHeader: true,});assert.deepEqual(webSessionCredentials.getWebSessionCredentialRequirement("copilot-web"), {