run-llama/llama_index · error · NotImplementedError

prefix_messages is not supported for CondenseQuestionChatEng

Error message

prefix_messages is not supported for CondenseQuestionChatEngine.

What it means

CondenseQuestionChatEngine.from_defaults() raises NotImplementedError when prefix_messages is passed. The engine delegates each turn to a query engine after condensing the question, and never prepends prefix messages to LLM calls, so the parameter is explicitly unsupported rather than ignored.

Source

Thrown at llama-index-core/llama_index/core/chat_engine/condense_question.py:104

        llm: Optional[LLM] = None,
        **kwargs: Any,
    ) -> "CondenseQuestionChatEngine":
        """Initialize a CondenseQuestionChatEngine from default parameters."""
        condense_question_prompt = condense_question_prompt or DEFAULT_PROMPT

        llm = llm or Settings.llm

        chat_history = chat_history or []
        memory = memory or memory_cls.from_defaults(
            chat_history=chat_history, token_limit=llm.metadata.context_window - 256
        )

        if system_prompt is not None:
            raise NotImplementedError(
                "system_prompt is not supported for CondenseQuestionChatEngine."
            )
        if prefix_messages is not None:
            raise NotImplementedError(
                "prefix_messages is not supported for CondenseQuestionChatEngine."
            )

        return cls(
            query_engine,
            condense_question_prompt,
            memory,
            llm,
            verbose=verbose,
            callback_manager=Settings.callback_manager,
        )

    def _condense_question(
        self, chat_history: List[ChatMessage], last_message: str
    ) -> str:
        """
        Generate standalone question from conversation context and last message.
        """

View on GitHub (pinned to afd0fef371)

Solutions

1. Remove prefix_messages from the call for this engine type
2. Use ContextChatEngine or SimpleChatEngine, which accept prefix_messages
3. Use CondensePlusContextChatEngine with system_prompt for persona + context needs
4. If you control the condense prompt, encode necessary instructions there instead

Example fix

# before
engine = CondenseQuestionChatEngine.from_defaults(
    query_engine=qe,
    prefix_messages=[ChatMessage(role='system', content='Be terse')],
)  # NotImplementedError

# after
from llama_index.core.chat_engine import SimpleChatEngine
engine = SimpleChatEngine.from_defaults(
    prefix_messages=[ChatMessage(role='system', content='Be terse')],
)

if engine_cls is CondenseQuestionChatEngine:
    assert prefix_messages is None, 'CondenseQuestionChatEngine does not accept prefix_messages; use ContextChatEngine or SimpleChatEngine'

try:
    engine = CondenseQuestionChatEngine.from_defaults(query_engine=qe, **kwargs)
except NotImplementedError as e:
    raise ConfigError(str(e)) from e

When it happens

Trigger: CondenseQuestionChatEngine.from_defaults(query_engine=qe, prefix_messages=[ChatMessage(...)]) — the `if prefix_messages is not None` guard fires immediately after the system_prompt check. Typically reached via generic engine-construction code that forwards prefix_messages to every engine type.

Common situations: Factory code shared across chat engines passing prefix_messages uniformly; porting a SimpleChatEngine setup (which accepts prefix_messages) to CondenseQuestionChatEngine.

Related errors

• system_prompt is not supported for CondenseQuestionChatEngin
• Streaming is not enabled. Please use chat() instead.
• Streaming is not enabled. Please use achat() instead.
• Cannot specify both system_prompt and prefix_messages
• Cannot specify both system_prompt and prefix_messages

AI-assisted analysis of run-llama/llama_index@afd0fef371 (2026-08-15). Data as JSON: /api/errors/7383d61b1cbc3c89. Report an issue: GitHub.