crewAIInc/crewAI · error · RuntimeError

FirecrawlApp not properly initialized

Error message

FirecrawlApp not properly initialized

What it means

RuntimeError raised at the top of FirecrawlScrapeWebsiteTool._run when self._firecrawl is falsy — the FirecrawlApp client attribute was never populated. __init__ always assigns it on the success path, so this indicates the tool object reached _run without a completed initialization.

Source

Thrown at lib/crewai-tools/src/crewai_tools/tools/firecrawl_scrape_website_tool/firecrawl_scrape_website_tool.py:109

            if click.confirm(
                "You are missing the 'firecrawl-py' package. Would you like to install it?"
            ):
                import subprocess

                subprocess.run(["uv", "add", "firecrawl-py"], check=True)  # noqa: S607
                from firecrawl import (
                    FirecrawlApp,
                )
            else:
                raise ImportError(
                    "`firecrawl-py` package not found, please run `uv add firecrawl-py`"
                ) from None

        self._firecrawl = FirecrawlApp(api_key=api_key)

    def _run(self, url: str) -> Any:
        if not self._firecrawl:
            raise RuntimeError("FirecrawlApp not properly initialized")

        url = validate_url(url)
        return self._firecrawl.scrape(url=url, **self.config)


try:
    from firecrawl import FirecrawlApp  # noqa: F401

    if not getattr(FirecrawlScrapeWebsiteTool, "_model_rebuilt", False):
        FirecrawlScrapeWebsiteTool.model_rebuild()
        FirecrawlScrapeWebsiteTool._model_rebuilt = True  # type: ignore[attr-defined]
except ImportError:
    pass

View on GitHub (pinned to 754d7323be)

Solutions

1. Create the tool through its constructor with a valid API key.
2. In tests, set tool._firecrawl to a mock before calling _run.
3. Let __init__ failures propagate instead of retaining half-initialized tools.

Example fix

# before
tool._firecrawl = None
tool._run('https://x.com')  # RuntimeError

# after
tool = FirecrawlScrapeWebsiteTool(api_key=FIRECRAWL_API_KEY)
tool._run('https://x.com')

if not getattr(tool, '_firecrawl', None):
    raise RuntimeError('scrape tool lacks FirecrawlApp client')

def scrape_tool_ready(tool) -> bool:
    return bool(getattr(tool, '_firecrawl', None))

try:
    out = tool._run(url)
except RuntimeError as e:
    if 'not properly initialized' in str(e):
        raise  # rebuild the tool properly instead of retrying the broken instance

When it happens

Trigger: Calling _run(url) on a tool built via __new__/model_copy or a test double that never set _firecrawl; running an instance whose __init__ raised midway and the error was swallowed.

Common situations: Mocked unit tests that forget to inject a client; Pydantic rebuild/copy paths that skip custom __init__; broad except blocks masking init failures.

Related errors

• FirecrawlApp not properly initialized
• FirecrawlApp not properly initialized
• Client not initialized
• Client is not initialized
• Failed to initialize MCP Adapter: {e}

AI-assisted analysis of crewAIInc/crewAI@754d7323be (2026-08-15). Data as JSON: /api/errors/a25578d9e1f33717. Report an issue: GitHub.