Merge remote-tracking branch 'origin/main' into lforst/dum-e/edinburgh-c4d8064dec

# Conflicts:
#	e2e/scenarios/flue-instrumentation/versions/v1/package.json
#	e2e/scenarios/flue-instrumentation/versions/v1/pnpm-lock.yaml

Author: lforst

.agents/skills/instrumentation/SKILL.md

@@ -27,7 +27,7 @@ Map the change before editing:
 - Inputs are untrusted: treat args, results, events, headers, and metadata as hostile. Prototype pollution is a concrete risk here. Avoid unsafe property access patterns, prototype-sensitive operations, and unnecessary mutation of third-party objects.
 - Support both auto-instrumentation and manual instrumentation. Auto-instrumentation does not cover every environment, loader, or framework.
 - For orchestrion auto-instrumentation, prefer targeting public API functions. Instrumenting internal helpers is more likely to break across library versions.
-- Auto and manual paths should share logic. Prefer both paths emitting the same tracing-channel events, with provider plugins converting those events into spans/logs/errors. Manual wrappers should not directly emit observability data.
+- Auto and manual paths should share logic through the same typed channel. For new and migrated instrumentation, prefer `invoke` in manual wrappers and `intercept` in provider plugins so the target can be scoped with `AsyncLocalStorage.run()` and its arguments, receiver, or output can be patched. Keep tracing-style hooks only as the compatibility path for instrumentation that has not migrated yet. Manual wrappers should not directly emit observability data.
 - Reuse shared repo utilities before introducing local helpers. Check `js/util/index.ts`, neighboring instrumentation files, and existing plugins/wrappers for utilities like `isObject`, merge helpers, and sanitizers before adding ad hoc replacements.
 - If a public instrumentation surface changes, check whether the export surface also needs updates in `js/src/instrumentation/index.ts` or `js/src/exports.ts`.
 - Preserve async context propagation. Changes around tracing channels, stream patching, or loader hooks must keep the current span context across awaits and stream consumption.

.changeset/add-flue-v2-support.md

@@ -0,0 +1,5 @@
+---
+"braintrust": patch
+---
+
+feat: Add support for flue v2

.changeset/simple-orchestrion-wrapper.md

@@ -0,0 +1,5 @@
+---
+"braintrust": patch
+---
+
+ref: Add simple wrapper to orchestrion transform

.changeset/vitest-evals-direct-input.md

@@ -0,0 +1,5 @@
+---
+"braintrust": minor
+---
+
+feat(vitest-evals): Add `meta.eval.input`

AGENTS.md

@@ -31,6 +31,8 @@ Use the normal Orchestrion config plus plugin/channel path by default. Special-c
 
 Instrumentation patches generally do not need to be removed during teardown. Prefer leaving behavior-preserving patches installed when they are idempotent; do not add unpatching machinery by default.
 
+Span names should generally remain stable across calls and versions. Do not include dynamic values such as model names in span names; record those values in metadata instead.
+
 ## Testing
 
 Uses Vitest. Prefer running the **narrowest relevant test** rather than the full suite.

e2e/README.md

@@ -121,6 +121,8 @@ Scenario-local manifests are optional and should stay slim. They are only for sc
 
 Scenarios that should participate in `test:e2e:bump` declare `braintrustScenario.bump.dependencies` in their existing scenario-local `package.json`. Each rule points at the real npm package and the tested range for that dependency alias; prereleases are ignored unless the rule sets `allowPrerelease: true`.
 
+Packages such as Next.js must be installed under their real dependency name. For a latest lane stored in a nested version manifest, set `targetManifest` to that scenario-relative `package.json` and `targetDependency` to the real dependency name. The bump script updates that manifest and regenerates its colocated lockfile while the rule key remains the logical `*-latest` lane name.
+
 `workspace:` dependency specs are intentionally not supported in scenario-local manifests. If a scenario needs a workspace package, keep that dependency in `e2e/package.json`.
 
 ## Running

e2e/config/pr-comment-scenarios.json

@@ -572,11 +572,27 @@
     "variants": [
       {
         "variantKey": "flue-v0-8-0",
-        "label": "v0.8.0"
+        "label": "v0.8 pinned"
+      },
+      {
+        "variantKey": "flue-v0-8-latest",
+        "label": "v0.8 latest"
       },
       {
         "variantKey": "flue-v1-0-0-beta-3",
-        "label": "v1.0.0-beta.3"
+        "label": "v1 pinned"
+      },
+      {
+        "variantKey": "flue-v1-latest",
+        "label": "v1 latest"
+      },
+      {
+        "variantKey": "flue-v2",
+        "label": "v2 pinned"
+      },
+      {
+        "variantKey": "flue-v2-latest",
+        "label": "v2 latest"
       }
     ]
   },
@@ -612,6 +628,36 @@
       }
     ]
   },
+  {
+    "scenarioDirName": "nextjs-instrumentation",
+    "label": "Next.js Instrumentation",
+    "metadataScenario": "nextjs-instrumentation",
+    "variants": [
+      {
+        "variantKey": "nextjs-v14",
+        "label": "v14 pinned"
+      },
+      {
+        "variantKey": "nextjs-v14-latest",
+        "label": "v14 latest"
+      }
+    ]
+  },
+  {
+    "scenarioDirName": "test-framework-evals-jest",
+    "label": "Jest Test Framework Evals",
+    "metadataScenario": "test-framework-evals-jest",
+    "variants": [
+      {
+        "variantKey": "jest-v29",
+        "label": "v29 pinned"
+      },
+      {
+        "variantKey": "jest-v29-latest",
+        "label": "v29 latest"
+      }
+    ]
+  },
   {
     "scenarioDirName": "test-framework-evals-vitest",
     "label": "Vitest Test Framework Evals",

e2e/helpers/normalize.test.ts

@@ -0,0 +1,12 @@
+import { describe, expect, test } from "vitest";
+import { normalizeForSnapshot } from "./normalize";
+
+describe("normalizeForSnapshot", () => {
+  test("normalizes a foreign repository path ending at a workspace directory", () => {
+    expect(
+      normalizeForSnapshot(
+        "find /another/checkout/braintrust-sdk-javascript/e2e -name AGENTS.md",
+      ),
+    ).toBe("find <repo>/e2e -name AGENTS.md");
+  });
+});

e2e/helpers/normalize.ts

@@ -110,7 +110,7 @@ const REPO_ROOT = path.resolve(HELPERS_DIR, "../..").replace(/\\/g, "/");
 const STACK_FRAME_REPO_PATH_REGEX =
   /(?:[A-Za-z]:)?[^\s)\n]*braintrust-sdk-javascript(?:[\\/](?:braintrust-sdk-javascript|[^\\/\s)\n]+))?((?:[\\/](?:e2e|js)[^:\s)\n]+)):\d+:\d+/g;
 const REPO_PATH_REGEX =
-  /(?:[A-Za-z]:)?[^\s)\n]*braintrust-sdk-javascript(?:[\\/](?:braintrust-sdk-javascript|[^\\/\s)\n]+))?((?:[\\/](?:e2e|js)[^:\s)\n]+))/g;
+  /(?:[A-Za-z]:)?[^\s)\n]*braintrust-sdk-javascript(?:[\\/](?:braintrust-sdk-javascript|[^\\/\s)\n]+))?((?:[\\/](?:e2e|js)[^:\s)\n]*))/g;
 const NODE_INTERNAL_FRAME_REGEX = /node:[^)\n]+:\d+:\d+/g;
 const TEMP_SCENARIO_PATH_REGEX =
   /\/e2e\/\.bt-tmp\/[^/\s)]+\/scenarios\/([^/\s)]+)\/?/g;