microsoft/autogen · error · ValueError

JSON content must be a dict

Error message

JSON content must be a dict

What it means

ChromaDBVectorMemory's internal text converter (used to turn MemoryContent into text for embedding/storage) accepts JSON mime type only when the inner content is a dict. A JSON content that is a list, string, or scalar raises this ValueError before the item is embedded.

Source

Thrown at python/packages/autogen-ext/src/autogen_ext/memory/chromadb/_chromadb.py:298

            except Exception as e:
                logger.error(f"Failed to get/create collection: {e}")
                raise

    def _extract_text(self, content_item: str | MemoryContent) -> str:
        """Extract searchable text from content."""
        if isinstance(content_item, str):
            return content_item

        content = content_item.content
        mime_type = content_item.mime_type

        if mime_type in [MemoryMimeType.TEXT, MemoryMimeType.MARKDOWN]:
            return str(content)
        elif mime_type == MemoryMimeType.JSON:
            if isinstance(content, dict):
                # Store original JSON string representation
                return str(content).lower()
            raise ValueError("JSON content must be a dict")
        elif isinstance(content, Image):
            raise ValueError("Image content cannot be converted to text")
        else:
            raise ValueError(f"Unsupported content type: {mime_type}")

    def _calculate_score(self, distance: float) -> float:
        """Convert ChromaDB distance to a similarity score."""
        if self._config.distance_metric == "cosine":
            return 1.0 - (distance / 2.0)
        return 1.0 / (1.0 + distance)

    async def update_context(
        self,
        model_context: ChatCompletionContext,
    ) -> UpdateContextResult:
        messages = await model_context.get_messages()
        if not messages:
            return UpdateContextResult(memories=MemoryQueryResult(results=[]))

View on GitHub (pinned to 027ecf0a37)

Solutions

1. Wrap list/scalar payloads in a dict before creating MemoryContent.
2. Use MemoryMimeType.TEXT for already-serialized JSON strings.
3. Validate content shape at the point where MemoryContent objects are constructed.

Example fix

# before
await memory.add(MemoryContent(content="[1, 2, 3]", mime_type=MemoryMimeType.JSON))
# after
await memory.add(MemoryContent(content={"items": [1, 2, 3]}, mime_type=MemoryMimeType.JSON))

from autogen_core.memory import MemoryContent, MemoryMimeType
if c.mime_type == MemoryMimeType.JSON:
    assert isinstance(c.content, dict), "ChromaDB memory JSON content must be a dict"
await memory.add(c)

def is_valid_json_memory_content(c) -> bool:
    if c.mime_type != MemoryMimeType.JSON:
        return True
    return isinstance(c.content, dict)

When it happens

Trigger: Calling memory.add() / update_context() with MemoryContent(content=<list or str>, mime_type=MemoryMimeType.JSON); the converter runs during ingestion and rejects non-dict JSON.

Common situations: Storing JSON arrays or pre-stringified JSON in ChromaDB memory; upstream producers emitting list payloads; reusing the same content objects across memory backends with mismatched expectations.

Related errors

• JSON content must be a dict
• Image content cannot be converted to text
• Unsupported content type: {mime_type}
• Image content cannot be converted to text
• Unsupported content type: {mime_type}

AI-assisted analysis of microsoft/autogen@027ecf0a37 (2026-08-15). Data as JSON: /api/errors/e19dc91ad3cc3a87. Report an issue: GitHub.