oraios/serena · error · ValueError

Content for {memory_name} is too long. Max length is {max_ch

Error message

Content for {memory_name} is too long. Max length is {max_chars} characters. Please make the content shorter.

What it means

The create_memory tool enforces a maximum character length on memory content before persisting it via MemoriesManager. If the content exceeds max_chars (or the configured default_max_tool_answer_chars when max_chars=-1), it refuses to save and asks for shorter content.

Source

Thrown at src/serena/tools/memory_tools.py:31

    """

    def apply(self, memory_name: str, content: str, max_chars: int = -1) -> str:
        """
        Write information about this project that can be useful for future tasks in md format.
        The name should be meaningful and can include "/" to organize into topics.
        If explicitly instructed, use the "global/" prefix for writing a memory that is shared across projects.
        References to other memories should be inside backticks and prefixed with mem:,
        e.g., `mem:auth`.

        :param memory_name: memory name
        :param content: memory content, utf8-encoded
        :param max_chars: see other tools
        """
        # NOTE: utf-8 encoding is configured in the MemoriesManager
        if max_chars == -1:
            max_chars = self.agent.serena_config.default_max_tool_answer_chars
        if len(content) > max_chars:
            raise ValueError(
                f"Content for {memory_name} is too long. Max length is {max_chars} characters. " + "Please make the content shorter."
            )

        return self.memory_manager.save_memory(memory_name, content, is_tool_context=True)


class ReadMemoryTool(Tool):
    """
    Reads the content of a memory file.
    """

    def apply(self, memory_name: str) -> str:
        """
        Use to read a memory that is likely to be relevant to the current task, inferring relevance e.g. from the name.
        """
        return self.memory_manager.load_memory(memory_name)

View on GitHub (pinned to 7fcbca7e62)

Solutions

1. Shorten/summarize the content below max_chars before saving
2. Split the content into multiple smaller memories
3. Pass a larger explicit max_chars value in the tool call
4. Raise serena_config.default_max_tool_answer_chars if -1 default is the binding limit

Example fix

// before
create_memory.apply(memory_name='notes', content=whole_file_text)  # 200k chars
// ValueError: Content too long...

// after
summary = summarize(whole_file_text)[:max_chars]
create_memory.apply(memory_name='notes', content=summary)
# or
create_memory.apply(memory_name='notes', content=whole_file_text, max_chars=500000)

MAX = max_chars if max_chars != -1 else agent.serena_config.default_max_tool_answer_chars
if len(content) > MAX:
    content = content[:MAX]  # or split/summarize
memory_tool.apply(memory_name=name, content=content)

try:
    memory_tool.apply(memory_name=name, content=content)
except ValueError as e:
    if 'too long' in str(e):
        memory_tool.apply(memory_name=name, content=summarize(content))
    else:
        raise

When it happens

Trigger: Calling the memory-creation tool with content whose len(content) exceeds max_chars; passing max_chars=-1 while the content is longer than serena_config.default_max_tool_answer_chars.

Common situations: An agent tries to save a whole file dump or a long analysis as a memory; default answer-length limit is low and the caller never raised it; content grew after a refactor and silently crossed the cap.

Related errors

• Bare "{self.GLOBAL_TOPIC}" is not a valid memory name. Use "
• Must be positive or the default (-1), got: {max_answer_chars
• Cannot use both fixed_tools and excluded_tools/included_opti
• Unknown language backend '{backend_str}': valid values are {
• Invalid line_ending: {value!r}. Valid values are: {valid}

AI-assisted analysis of oraios/serena@7fcbca7e62 (2026-08-29). Data as JSON: /api/errors/979ff9e0c895681f. Report an issue: GitHub.