BlockNote AI table generation is unstable; long tables fail with ChunkExecutionError
@YousefED ya está trabajando en esto.
Desde el 16/2/2026.
Evaluación
Este issue todavía no se ha evaluado.
Descripción
Description
While using BlockNote AI to generate tables, I encountered two related issues.
Both issues are reproducible in the official AI Playground.
AI Playground:
https://www.blocknotejs.org/examples/ai/playground
Versions
{
"@blocknote/core": "^0.45.0",
"@blocknote/mantine": "^0.45.0",
"@blocknote/react": "^0.45.0",
"@blocknote/xl-ai": "^0.45.0"
}
Issue 1: Table generation is unreliable
Whenever the prompt asks the AI to generate a table, there is a non-deterministic failure rate.
Even for short tables, the AI sometimes fails immediately and shows an error instead of producing a table block.
Issue 2: Long tables fail during generation
When asking the AI to generate larger or longer tables (e.g. more rows or more content per cell), the generation process is interrupted and consistently fails.
Observed Error
ChunkExecutionError: Tool execution failed: Slice has openStart or openEnd > 0, but structure=false
Partial error object
{
result: {
error: ChunkExecutionError: Tool execution failed: Slice has openStart or openEnd > 0, but structure=false
},
chatStatus: "ready",
chatError: undefined
}
The stack trace points to BlockNote’s internal transform / execute logic, suggesting a structural issue during streaming or chunk application.
Steps to Reproduce
- Open the official BlockNote AI Playground
- Enter a prompt asking the AI to generate a table
- Short table (few rows): sometimes succeeds, sometimes errors
- Long table (many rows / larger content): generation consistently fails mid-process
- The table block is not fully created
Expected Behavior
- AI table generation should be stable or at least retryable
- Long tables should not break the ProseMirror document structure during streaming
Actual Behavior
- Table generation is unstable
- Long tables fail with a
ChunkExecutionErrorand interrupt the generation process
Additional Notes
- Manual table insertion works correctly
- The issue only occurs when using BlockNote AI (
@blocknote/xl-ai) - This appears to be related to streaming chunk handling and incomplete table slice structure (
structure=false)
- Lenguaje dominante
- TypeScript
- Estrellas
- 10.2k
- Forks
- 772
- Merge medio
- 3 d 11 h
- PR fusionados (30 d)
- 17
Guía de contribución
Primeros pasos
- Lee el issue completo y luego la guía de contribución del proyecto.
- Comenta en el issue que vas a ocuparte — evita que dos personas hagan lo mismo.
- Haz un fork del repositorio y trabaja en una rama.
- Abre un pull request que haga referencia al número del issue.
Más de TypeCellOS/BlockNote
-
Dificultad 2/5 1-3 horas Aptitud para principiantes 84/100
TypeCellOS/BlockNote#3098 ·
-
Off-By-One `RangeError` Crash and Text Duplication in `StyleManager.editLink` and `deleteLink` Abiertoneeds-triage
Dificultad 2/5 1-3 horas Aptitud para principiantes 78/100
TypeCellOS/BlockNote#3073 ·
-
needs-triage
Dificultad 2/5 1-3 horas Aptitud para principiantes 85/100
TypeCellOS/BlockNote#3072 ·
-
Dificultad 2/5 1-3 horas Aptitud para principiantes 82/100
TypeCellOS/BlockNote#2949 · 1 comentario ·
-
a11y
Dificultad 2/5 1-3 horas Aptitud para principiantes 68/100
TypeCellOS/BlockNote#2855 ·
Todos los issues de TypeCellOS/BlockNote
Issues similares
-
clawsweeper:fix-shape-clear clawsweeper:queueable-fix clawsweeper:source-repro impact:ux-friction issue-rating: 🦞 diamond lobster no-stale P3
Dificultad 2/5 1-3 horas Aptitud para principiantes 78/100
-
community first-timers-only good first issue hacktoberfest help wanted low hanging fruit up-for-grabs
Dificultad 1/5 Menos de una hora Aptitud para principiantes 76/100
-
code-quality refactoring
Dificultad 2/5 1-3 horas Aptitud para principiantes 84/100
github/gh-aw-firewall#8816 ·
-
integration:quickjs org:external priority:backlog topic:code-interpreter topic:middleware type:feature
Dificultad 2/5 1-3 horas Aptitud para principiantes 74/100
langchain-ai/deepagents#6450 ·
-
Dificultad 1/5 Menos de una hora Aptitud para principiantes 88/100
vercel/react-tweet#225 ·