Merge origin/main into step-dispatch-run-context

Author: TooTallNate

.changeset/atomic-postgres-run-creation.md

@@ -0,0 +1,5 @@
+---
+'@workflow/world-postgres': patch
+---
+
+Fix run creation and slot allocation not happening transactionally

.changeset/batch-pair-chunk.md

@@ -0,0 +1,5 @@
+---
+"@workflow/core": patch
+---
+
+Commit the batched fan-out's pre-claimed inline `[step_created, step_started]` pairs in their own leading chunk, ahead of the plain step/wait creates, so the write that gates the inline bodies stays small and commits faster. The pairs now fold only with two or more inline steps (a lone inline step keeps the lazy start), and a lone plain create beside them takes the single write.

.changeset/breezy-days-run.md

@@ -0,0 +1,4 @@
+---
+---
+
+Persist a bounded 30-run history of E2E retry rates with exact lane dimensions and execution denominators.

.changeset/chubby-bears-check.md

@@ -0,0 +1,4 @@
+---
+---
+
+Give Windows cancellation E2E tests the standard CLI timeout budget.

.changeset/disable-rollup-input-maps.md

@@ -0,0 +1,5 @@
+---
+"@workflow/rollup": patch
+---
+
+Disable implicit input source map loading during workflow transforms to avoid false missing-map build errors for dependencies.

.changeset/fluffy-apes-worry.md

@@ -0,0 +1,7 @@
+---
+"workflow": patch
+"@workflow/core": patch
+"@workflow/ai": patch
+---
+
+Route Workflow AI examples through AI Gateway and recommend WorkflowAgent for Workflow 5.

.changeset/gateway-classid-redefine-crash.md

@@ -0,0 +1,5 @@
+---
+'@workflow/swc-plugin': patch
+---
+
+Fix a crash ("Cannot redefine property: classId") when a bundler pipeline re-runs the transform over its own output for a dependency that ships custom serialization methods, such as `@ai-sdk/gateway`.

.changeset/postgres-skipped-stream-cursor.md

@@ -0,0 +1,5 @@
+---
+'@workflow/world-postgres': patch
+---
+
+Fix stream cursors not being advanced for skipped chunks, so repeated notifications cannot consume the requested start offset twice

.changeset/quickjs-slot-report-parity.md

@@ -0,0 +1,5 @@
+---
+'@workflow/core': patch
+---
+
+The QuickJS engine now reports its log position on writes and feeds its VM from the events a World returns on the response, matching the node:vm engine.

.github/scripts/aggregate-e2e-results.js

@@ -2,6 +2,7 @@
 
 const fs = require('node:fs');
 const path = require('node:path');
+const { dimensionFor } = require('./generate-e2e-flake-history.js');
 
 // Parse command line arguments
 const args = process.argv.slice(2);
@@ -62,7 +63,7 @@ function findResultFiles(dir) {
     // trailing dot matters — the Python lane's report is
     // `e2e-conformance-python.json` and must keep matching.
     'e2e-conformance.',
-  ]);
+  ]).filter((file) => !file.endsWith('.flaky.json'));
 }
 
 // Find all e2e metadata JSON files
@@ -159,19 +160,34 @@ function loadFailures(dir) {
 // several jobs for one app is collapsed into a single entry with an
 // occurrence count so the section stays scannable.
 function loadFlaky(dir) {
-  // Map of `${app}\u0000${testName}` -> { app, testName, retryCount, occurrences }
+  // New sidecars pair 1:1 with exact result files; keep reading legacy names
+  // while artifacts from older branches can still reach the aggregator.
   const flaky = new Map();
-  const files = findJsonFiles(dir, 'e2e-flaky-');
-
-  for (const file of files) {
-    const basename = path.basename(file, '.json');
-    const match = basename.match(/^e2e-flaky-(.+)-(?:vercel|local)$/);
-    const app = match ? match[1] : 'unknown';
+  const paired = findJsonFiles(dir, 'e2e-').filter((file) =>
+    file.endsWith('.flaky.json')
+  );
+  const legacy = findJsonFiles(dir, 'e2e-flaky-');
+
+  for (const file of [...paired, ...legacy]) {
+    const basename = path.basename(file);
+    const pairedReport = basename.replace(/\.flaky\.json$/, '.json');
+    const dimension = basename.endsWith('.flaky.json')
+      ? dimensionFor(pairedReport)
+      : null;
+    const legacyMatch = path
+      .basename(file, '.json')
+      .match(/^e2e-flaky-(.+)-(?:vercel|local)$/);
+    const app = dimension?.app || legacyMatch?.[1] || 'unknown';
+    const lane = dimension
+      ? [dimension.lane, dimension.world, dimension.vm, dimension.variant]
+          .filter(Boolean)
+          .join(' / ')
+      : null;
     try {
       const entries = JSON.parse(fs.readFileSync(file, 'utf-8'));
       for (const entry of entries) {
         if (!entry.testName) continue;
-        const key = `${app}\u0000${entry.testName}`;
+        const key = `${lane || app}\u0000${entry.fullName || entry.testName}`;
         const existing = flaky.get(key);
         if (existing) {
           existing.occurrences++;
@@ -182,6 +198,7 @@ function loadFlaky(dir) {
         } else {
           flaky.set(key, {
             app,
+            lane,
             testName: entry.testName,
             retryCount: entry.retryCount || 1,
             occurrences: 1,
@@ -286,7 +303,8 @@ function renderFlakySection(flakyTests) {
   for (const test of sorted) {
     const jobs =
       test.occurrences > 1 ? ` — flaked in ${test.occurrences} jobs` : '';
-    console.log(`- \`${test.testName}\` (${test.app})${jobs}`);
+    const location = test.lane ? `${test.app} · ${test.lane}` : test.app;
+    console.log(`- \`${test.testName}\` (${location})${jobs}`);
   }
   console.log('');
   if (collapse) {