[ai] Fix "missing text part" stream crash by repairing UI message framing#2537
Draft
VaguelySerious wants to merge 1 commit into
Draft
[ai] Fix "missing text part" stream crash by repairing UI message framing#2537VaguelySerious wants to merge 1 commit into
VaguelySerious wants to merge 1 commit into
Conversation
…ming Duplicated or interleaved durable stream writes (step retry/redelivery or concurrent-worker duplication) can land a finish-step in the middle of a text part that reuses id "0", orphaning the rest. The AI SDK consumer then throws `Received text-delta for missing text part with ID "0"` and kills the turn. WorkflowChatTransport now normalizes UI message stream part framing — mirroring the consumer's per-step part-lifetime state machine — so orphaned deltas/ends get a synthesized start and replayed chunks are dropped. Well-formed streams pass through unchanged. Closes #2422 Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
🦋 Changeset detectedLatest commit: f4f35c5 The changes in this PR will be included in the next version bump. This PR includes changesets to release 1 package
Not sure what this means? Click here to learn what changesets are. Click here if you're a maintainer who wants to add another changeset to this PR |
Contributor
Contributor
🧪 E2E Test Results✅ All tests passed Summary
Details by Category✅ ▲ Vercel Production
✅ 💻 Local Development
✅ 📦 Local Production
✅ 🐘 Local Postgres
✅ 🪟 Windows
✅ 📋 Other
|
Contributor
📊 Benchmark Results
workflow with no steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Nitro | Express | Next.js (Turbopack) workflow with 1 step💻 Local Development
▲ Production (Vercel)
🔍 Observability: Next.js (Turbopack) | Express | Nitro workflow with 10 sequential steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Next.js (Turbopack) | Express | Nitro workflow with 25 sequential steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Nitro | Express | Next.js (Turbopack) workflow with 50 sequential steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Nitro | Next.js (Turbopack) Promise.all with 10 concurrent steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Next.js (Turbopack) | Nitro Promise.all with 25 concurrent steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Nitro | Next.js (Turbopack) | Express Promise.all with 50 concurrent steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Nitro | Express | Next.js (Turbopack) Promise.race with 10 concurrent steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Next.js (Turbopack) | Express | Nitro Promise.race with 25 concurrent steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Nitro | Express | Next.js (Turbopack) Promise.race with 50 concurrent steps💻 Local Development
▲ Production (Vercel)
🔍 Observability: Nitro | Next.js (Turbopack) | Express workflow with 10 sequential data payload steps (10KB)💻 Local Development
▲ Production (Vercel)
🔍 Observability: Nitro | Express | Next.js (Turbopack) workflow with 25 sequential data payload steps (10KB)💻 Local Development
▲ Production (Vercel)
🔍 Observability: Nitro | Express | Next.js (Turbopack) workflow with 50 sequential data payload steps (10KB)💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Next.js (Turbopack) | Nitro workflow with 10 concurrent data payload steps (10KB)💻 Local Development
▲ Production (Vercel)
🔍 Observability: Nitro | Express | Next.js (Turbopack) workflow with 25 concurrent data payload steps (10KB)💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Nitro | Next.js (Turbopack) workflow with 50 concurrent data payload steps (10KB)💻 Local Development
▲ Production (Vercel)
🔍 Observability: Nitro | Express | Next.js (Turbopack) Stream Benchmarks (includes TTFB metrics)workflow with stream💻 Local Development
▲ Production (Vercel)
🔍 Observability: Next.js (Turbopack) | Express | Nitro stream pipeline with 5 transform steps (1MB)💻 Local Development
▲ Production (Vercel)
🔍 Observability: Nitro | Express | Next.js (Turbopack) 10 parallel streams (1MB each)💻 Local Development
▲ Production (Vercel)
🔍 Observability: Express | Nitro | Next.js (Turbopack) fan-out fan-in 10 streams (1MB each)💻 Local Development
▲ Production (Vercel)
🔍 Observability: Nitro | Express | Next.js (Turbopack) SummaryFastest Framework by WorldWinner determined by most benchmark wins
Fastest World by FrameworkWinner determined by most benchmark wins
Column Definitions
Worlds:
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Closes #2422
Problem
A workflow run owns a single shared UI message stream, and the AI SDK's stream consumer (
processUIMessageStream, backinguseChat/readUIMessageStream) resets its active text/reasoning part maps on everyfinish-step. Multi-step turns reuse the same part id (commonly"0") in each step. So if atext-startis dropped or afinish-steplands in the middle of a text part, the consumer hits atext-delta/text-endfor an id it has no open part for and throws:…which kills the whole turn.
That malformed framing is reachable whenever a stream-producing step's output is duplicated or interleaved on the shared stream — e.g. step retry/redelivery, reconnect/replay overlap, or the concurrent-worker duplication tracked in #2331 / #2039 (a
finish-stepfrom one execution landing inside another's text part).Fix
WorkflowChatTransportnow passes the stream it hands to the AI SDK throughnormalizeUIMessageStreamParts, which mirrors the consumer's per-step part-lifetime state machine:finish-step(exactly where the consumer resets);*-startwhen an orphaned*-delta/*-endarrives;*-start/*-delta/*-endfor a part already open/ended in the current step (reconnect/replay overlap).It wraps the output of the send/reconnect iterators, so the reconnect
chunkIndex/gotFinishaccounting is untouched. A well-formed stream passes through unchanged; the worst case degrades to "text begins slightly into the step" or "a duplicated tail is dropped" instead of a dead turn. Applies to both text and reasoning parts.This addresses the client-visible symptom for all
WorkflowChatTransportconsumers regardless of the upstream duplication cause. The underlying server-side duplicate-execution races (#2331, #2039) are separate and remain tracked there.Tests
packages/ai/src/workflow-chat-transport.stream-repair.test.tsdrives the real transport (mockfetchreturning chunks via the AI SDK's ownJsonToSseTransformStream) into the realreadUIMessageStreamconsumer:missing text parterror;"Hello world");"0"reused across steps) is unchanged.🤖 Generated with Claude Code