ci: reduce node runner fanout

.github/workflows/ci.yml

@@ -746,6 +746,7 @@ jobs:
     timeout-minutes: 60
     strategy:
       fail-fast: false
+      max-parallel: 3
       matrix: ${{ fromJson(needs.preflight.outputs.channel_contracts_matrix) }}
     steps:
       - name: Checkout
@@ -1118,6 +1119,7 @@ jobs:
     timeout-minutes: 60
     strategy:
       fail-fast: false
+      max-parallel: 8
       matrix: ${{ fromJson(needs.preflight.outputs.checks_node_core_nondist_matrix) }}
     steps:
       - name: Checkout
@@ -1355,6 +1357,7 @@ jobs:
     timeout-minutes: 20
     strategy:
       fail-fast: false
+      max-parallel: 3
       matrix:
         include:
           - check_name: check-preflight-guards
@@ -1495,6 +1498,7 @@ jobs:
     timeout-minutes: 20
     strategy:
       fail-fast: false
+      max-parallel: 3
       matrix:
         include:
           - check_name: check-additional-boundaries

docs/ci.md

@@ -62,7 +62,7 @@ Local changed-lane logic lives in `scripts/changed-lanes.mjs` and is executed by
 
 On pushes, the `checks` matrix adds the push-only `compat-node22` lane. On pull requests, that lane is skipped and the matrix stays focused on the normal test/channel lanes.
 
-The slowest Node test families are split or balanced so each job stays small: channel contracts split registry and core coverage into six weighted shards total, bundled plugin tests balance across six extension workers, auto-reply runs as three balanced workers instead of six tiny workers, and agentic gateway/plugin configs are spread across the existing source-only agentic Node jobs instead of waiting on built artifacts. Broad browser, QA, media, and miscellaneous plugin tests use their dedicated Vitest configs instead of the shared plugin catch-all. The broad agents lane uses the shared Vitest file-parallel scheduler because it is import/scheduling dominated rather than owned by a single slow test file. `runtime-config` runs with the infra core-runtime shard to keep the shared runtime shard from owning the tail. `check-additional` keeps package-boundary compile/canary work together and separates runtime topology architecture from gateway watch coverage; the boundary guard shard runs its small independent guards concurrently inside one job. Gateway watch, channel tests, and the core support-boundary shard run concurrently inside `build-artifacts` after `dist/` and `dist-runtime/` are already built, keeping their old check names as lightweight verifier jobs while avoiding two extra Blacksmith workers and a second artifact-consumer queue.
+The slowest Node test families are split or balanced so each job stays small without over-reserving runners: channel contracts run as three weighted shards, bundled plugin tests balance across six extension workers, small core unit lanes are paired, auto-reply runs as three balanced workers instead of six tiny workers, and agentic gateway/plugin configs are spread across the existing source-only agentic Node jobs instead of waiting on built artifacts. Broad browser, QA, media, and miscellaneous plugin tests use their dedicated Vitest configs instead of the shared plugin catch-all. The broad agents lane uses the shared Vitest file-parallel scheduler because it is import/scheduling dominated rather than owned by a single slow test file. `runtime-config` runs with the infra core-runtime shard to keep the shared runtime shard from owning the tail. `check-additional` keeps package-boundary compile/canary work together and separates runtime topology architecture from gateway watch coverage; the boundary guard shard runs its small independent guards concurrently inside one job. Gateway watch, channel tests, and the core support-boundary shard run concurrently inside `build-artifacts` after `dist/` and `dist-runtime/` are already built, keeping their old check names as lightweight verifier jobs while avoiding two extra Blacksmith workers and a second artifact-consumer queue.
 Android CI runs both `testPlayDebugUnitTest` and `testThirdPartyDebugUnitTest`, then builds the Play debug APK. The third-party flavor has no separate source set or manifest; its unit-test lane still compiles that flavor with the SMS/call-log BuildConfig flags, while avoiding a duplicate debug APK packaging job on every Android-relevant push.
 `extension-fast` is PR-only because push runs already execute the full bundled plugin shards. That keeps changed-plugin feedback for reviews without reserving an extra Blacksmith worker on `main` for coverage already present in `checks-node-extensions`.
 

scripts/lib/channel-contract-test-plan.mjs

@@ -41,14 +41,9 @@ export function createChannelContractTestShards() {
   const rootDir = "src/channels/plugins/contracts";
   const suffixes = ["a", "b", "c"];
   const groups = Object.fromEntries(
-    ["registry", "core"].flatMap((family) =>
-      suffixes.map((suffix) => [`checks-fast-contracts-channels-${family}-${suffix}`, []]),
-    ),
+    suffixes.map((suffix) => [`checks-fast-contracts-channels-${suffix}`, []]),
   );
-  const groupKeys = {
-    core: suffixes.map((suffix) => `checks-fast-contracts-channels-core-${suffix}`),
-    registry: suffixes.map((suffix) => `checks-fast-contracts-channels-registry-${suffix}`),
-  };
+  const groupKeys = suffixes.map((suffix) => `checks-fast-contracts-channels-${suffix}`);
   const weights = Object.fromEntries(Object.keys(groups).map((key) => [key, 0]));
   const pushBalanced = (keys, file) => {
     const target = keys.toSorted((a, b) => weights[a] - weights[b] || a.localeCompare(b))[0];
@@ -71,10 +66,10 @@ export function createChannelContractTestShards() {
     return delta === 0 ? left.localeCompare(right) : delta;
   };
   for (const file of registryFiles.toSorted(byDescendingWeight)) {
-    pushBalanced(groupKeys.registry, file);
+    pushBalanced(groupKeys, file);
   }
   for (const file of coreFiles.toSorted(byDescendingWeight)) {
-    pushBalanced(groupKeys.core, file);
+    pushBalanced(groupKeys, file);
   }
 
   return Object.entries(groups).map(([checkName, includePatterns]) => ({

scripts/lib/ci-node-test-plan.mjs

@@ -87,6 +87,36 @@ function createAutoReplyReplySplitShards() {
 }
 
 const SPLIT_NODE_SHARDS = new Map([
+  [
+    "core-unit-fast",
+    [
+      {
+        shardName: "core-unit-fast-support",
+        configs: [
+          "test/vitest/vitest.unit-fast.config.ts",
+          "test/vitest/vitest.unit-support.config.ts",
+        ],
+        includeExternalConfigs: true,
+        requiresDist: false,
+      },
+    ],
+  ],
+  [
+    "core-unit-src",
+    [
+      {
+        shardName: "core-unit-src-security",
+        configs: [
+          "test/vitest/vitest.unit-src.config.ts",
+          "test/vitest/vitest.unit-security.config.ts",
+        ],
+        includeExternalConfigs: true,
+        requiresDist: false,
+      },
+    ],
+  ],
+  ["core-unit-security", []],
+  ["core-unit-support", []],
   [
     "core-runtime",
     [
@@ -205,7 +235,9 @@ export function createNodeTestShards() {
     const splitShards = SPLIT_NODE_SHARDS.get(shard.name);
     if (splitShards) {
       return splitShards.flatMap((splitShard) => {
-        const splitConfigs = splitShard.configs.filter((config) => configs.includes(config));
+        const splitConfigs = splitShard.includeExternalConfigs
+          ? splitShard.configs
+          : splitShard.configs.filter((config) => configs.includes(config));
         if (splitConfigs.length === 0) {
           return [];
         }

test/scripts/channel-contract-test-plan.test.ts

@@ -25,13 +25,11 @@ describe("scripts/lib/channel-contract-test-plan.mjs", () => {
         task: shard.task,
       })),
     ).toEqual(
-      ["registry", "core"].flatMap((family) =>
-        suffixes.map((suffix) => ({
-          checkName: `checks-fast-contracts-channels-${family}-${suffix}`,
-          runtime: "node",
-          task: "contracts-channels",
-        })),
-      ),
+      suffixes.map((suffix) => ({
+        checkName: `checks-fast-contracts-channels-${suffix}`,
+        runtime: "node",
+        task: "contracts-channels",
+      })),
     );
   });
 
@@ -45,13 +43,11 @@ describe("scripts/lib/channel-contract-test-plan.mjs", () => {
   });
 
   it("keeps registry-backed surface shards spread across checks", () => {
-    for (const shard of createChannelContractTestShards().filter((entry) =>
-      entry.checkName.includes("-registry-"),
-    )) {
+    for (const shard of createChannelContractTestShards()) {
       const surfaceRegistryFiles = shard.includePatterns.filter((pattern) =>
         pattern.includes("/surfaces-only.registry-backed-shard-"),
       );
-      expect(surfaceRegistryFiles.length).toBeLessThanOrEqual(3);
+      expect(surfaceRegistryFiles.length).toBeLessThanOrEqual(4);
     }
   });
 });

test/scripts/ci-node-test-plan.test.ts

@@ -25,6 +25,40 @@ function listTestFiles(rootDir: string): string[] {
 }
 
 describe("scripts/lib/ci-node-test-plan.mjs", () => {
+  it("combines the small core unit shards to reduce CI runner fanout", () => {
+    const coreUnitShards = createNodeTestShards()
+      .filter((shard) => shard.shardName.startsWith("core-unit-"))
+      .map((shard) => ({
+        configs: shard.configs,
+        requiresDist: shard.requiresDist,
+        shardName: shard.shardName,
+      }));
+
+    expect(coreUnitShards).toEqual([
+      {
+        configs: [
+          "test/vitest/vitest.unit-fast.config.ts",
+          "test/vitest/vitest.unit-support.config.ts",
+        ],
+        requiresDist: false,
+        shardName: "core-unit-fast-support",
+      },
+      {
+        configs: [
+          "test/vitest/vitest.unit-src.config.ts",
+          "test/vitest/vitest.unit-security.config.ts",
+        ],
+        requiresDist: false,
+        shardName: "core-unit-src-security",
+      },
+      {
+        configs: ["test/vitest/vitest.unit-ui.config.ts"],
+        requiresDist: false,
+        shardName: "core-unit-ui",
+      },
+    ]);
+  });
+
   it("names the node shard checks as core test lanes", () => {
     const shards = createNodeTestShards();