crewAIInc/crewAI · error · FileNotFoundError

The following file does not exist: {source_content.source}

Error message

The following file does not exist: {source_content.source}

What it means

A plain FileNotFoundError raised by TextFileLoader.load when source_content.path_exists() returns False. It is the standard Python missing-file signal (not a ValueError), so existing except FileNotFoundError handling works. The message echoes source_content.source verbatim.

Source

Thrown at lib/crewai-tools/src/crewai_tools/rag/loaders/text_loader.py:11

from typing import Any

from crewai_tools.rag.base_loader import BaseLoader, LoaderResult
from crewai_tools.rag.source_content import SourceContent


class TextFileLoader(BaseLoader):
    def load(self, source_content: SourceContent, **kwargs: Any) -> LoaderResult:  # type: ignore[override]
        source_ref = source_content.source_ref
        if not source_content.path_exists():
            raise FileNotFoundError(
                f"The following file does not exist: {source_content.source}"
            )

        with open(source_content.source, encoding="utf-8") as file:
            content = file.read()

        return LoaderResult(
            content=content,
            source=source_ref,
            doc_id=self.generate_doc_id(source_ref=source_ref, content=content),
        )


class TextLoader(BaseLoader):
    def load(self, source_content: SourceContent, **kwargs: Any) -> LoaderResult:  # type: ignore[override]
        return LoaderResult(
            content=source_content.source,
            source=source_content.source_ref,

View on GitHub (pinned to 754d7323be)

Solutions

Check the path exists and resolve it to absolute before loading: os.path.isfile(os.path.abspath(path)).
If the path is relative, anchor it to a known base directory instead of relying on cwd.
Print repr(path) to expose trailing spaces, unicode look-alikes, or wrong case.
When the source string comes from LLM output, validate it against a whitelist directory first.

Example fix

# before
result = loader.load(SourceContent(path="notes.txt"))  # cwd-dependent

# after
import os
base = "/srv/data"
full = os.path.realpath(os.path.join(base, "notes.txt"))
if not os.path.isfile(full):
    raise SystemExit(f"missing file: {full}")
result = loader.load(SourceContent(path=full))

Defensive patterns

Strategy: validation

Validate before calling

import os

def resolve_input_file(path: str, base: str) -> str:
    full = os.path.realpath(os.path.join(base, path))
    if not os.path.isfile(full):
        raise FileNotFoundError(f"no such file: {full!r}")
    return full

Try / catch

try:
    result = text_loader.load(src)
except FileNotFoundError as e:
    log.info("skipping missing file: %s", src.source)

Prevention

Anchor relative paths to an explicit base directory, never cwd.
Use repr(path) in logs to expose whitespace and unicode issues.
Validate LLM-supplied filenames against an allowlisted directory before loading.

When it happens

Trigger: Calling text_loader.load(SourceContent(path='data/notes.txt')) when the file is absent, when the path is relative and the process cwd differs from what you assumed, or when the filename has trailing whitespace/invisible characters. Note the existence check uses path_exists() but open() uses source_content.source — a mismatched or URL-derived source can also reach open() with an invalid path and raise the OS-level FileNotFoundError instead.

Common situations: Running an agent from a different working directory than the shell where the path worked; paths passed by an LLM tool call that hallucinates filenames; case-sensitivity differences between macOS and Linux; files not committed/synced to the deployment environment.

Related errors

AI-assisted analysis of crewAIInc/crewAI@754d7323be (2026-08-15). Data as JSON: /api/errors/779433ac286c2056. Report an issue: GitHub.