python-poetry/poetry · error · RuntimeError

Package {link.url} cannot be installed in the current enviro

Error message

Package {link.url} cannot be installed in the current environment {self._env.marker_env}

What it means

Raised by Executor._download_link_archive at src/poetry/installation/executor.py:759-767 when get_cached_archive_for_link(link, strict=False, env=...) returns None. With strict=False the cache is allowed to fall back to the original link, so a None result means the original archive itself is not valid for the current environment (no compatible candidate). RuntimeError naming the URL and the marker env.

Source

Thrown at src/poetry/installation/executor.py:764

        # Get original package for the link provided
        download_func = functools.partial(self._download_archive, operation)
        original_archive = self._artifact_cache.get_cached_archive_for_link(
            link, strict=True, download_func=download_func
        )

        # Get potential higher prioritized cached archive, otherwise it will fall back
        # to the original archive.
        archive = self._artifact_cache.get_cached_archive_for_link(
            link,
            strict=False,
            env=self._env,
        )
        if archive is None:
            # Since we previously downloaded an archive, we now should have
            # something cached that we can use here. The only case in which
            # archive is None is if the original archive is not valid for the
            # current environment.
            raise RuntimeError(
                f"Package {link.url} cannot be installed in the current environment"
                f" {self._env.marker_env}"
            )

        if archive.suffix != ".whl":
            message = (
                f"  <fg=blue;options=bold>-</> {self.get_operation_message(operation)}:"
                f"{format_build_wheel_log(package, self._env)}"
            )
            self._write(operation, message)

            name = operation.package.name
            archive = self._chef.prepare(
                archive,
                output_dir=original_archive.parent,
                config_settings=self._build_config_settings.get(name),
                build_constraints=self._build_constraints.get(name),
            )

View on GitHub (pinned to 92b74dcfe3)

Solutions

Verify and switch to the interpreter the lock was resolved with: `poetry env info`, `poetry env use <python>`.
Regenerate the lock for the current env: `poetry lock --regenerate` then install.
Ensure an sdist is available so Poetry can build for the platform, or add a compatible source.
Check markers in pyproject.toml (python = "...") match the active interpreter.

Example fix

# before: lock made under py3.11, now in py3.9
$ poetry install
RuntimeError: Package ... cannot be installed in the current environment ...

# after
$ poetry env use python3.11
$ poetry install

Defensive patterns

Strategy: validation

Validate before calling

from poetry.utils.env import EnvManager

env = EnvManager.get_default_env()
for link in links:
    if link.is_wheel:
        from poetry.installation.wheel import Wheel
        if not Wheel(link.filename).is_supported_by_environment(env):
            continue
    # ensure at least one link is env-compatible before install
    break
else:
    raise ValueError('No artifact compatible with the current env')

Try / catch

try:
    executor.run(operations)
except RuntimeError as e:
    if 'cannot be installed in the current environment' in str(e):
        raise SystemExit('Use `poetry env use` to switch to the locked interpreter.') from e
    raise

Prevention

Resolve and lock under the same interpreter family you install with.
Keep python constraint in pyproject.toml aligned with the deployment interpreter.

When it happens

Trigger: Installing a package whose only artifacts (cached or freshly downloaded) are incompatible with the current env's markers — e.g. a wheel only built for a different Python/OS, or an sdist that needs a build the env cannot satisfy. Occurs during executor.run when materializing the archive.

Common situations: Locked dependency resolved for a different interpreter than the one currently active; a wheel-only package on an unsupported platform; env markers (python_version, sys_platform) excluding every artifact.

Related errors

AI-assisted analysis of python-poetry/poetry@92b74dcfe3 (2026-08-04). Data as JSON: /data/errors/278c2e10f7599be3.json. Report an issue: GitHub.