langchain-ai/deepagents · error · RuntimeError
interpreter.enable_interpreter is missing from the config ma
Error message
interpreter.enable_interpreter is missing from the config manifest
What it means
`_warn_if_interpreter_disabled_by_sandbox` raises `RuntimeError` if `get_option("interpreter.enable_interpreter")` returns `None`. The function needs the manifest-declared local default to decide whether to warn about sandbox suppression of the interpreter; without the manifest entry it cannot compute the default and fails instead of silently skipping the warning.
Source
Thrown at libs/code/deepagents_code/main.py:1480
it (the middleware is unsupported under a remote sandbox). This prints to
stderr on the non-interactive (`-n`) path; the interactive TUI surfaces the
same advisory as a startup notification (see
`DeepAgentsApp._notify_interpreter_disabled_by_sandbox`).
Keyed on the raw `args.interpreter` tri-state so an explicit
`--no-interpreter` opt-out stays silent (the predicate only fires for the
unset default).
Raises:
RuntimeError: If the interpreter option is absent from the manifest.
"""
from deepagents_code._server_config import _interpreter_suppressed_by_sandbox
from deepagents_code.config_manifest import get_option
option = get_option("interpreter.enable_interpreter")
if option is None:
msg = "interpreter.enable_interpreter is missing from the config manifest"
raise RuntimeError(msg)
local_default = bool(_resolver_for_args(args).get(option).value)
if not _interpreter_suppressed_by_sandbox(
enable_interpreter=args.interpreter,
sandbox_type=args.sandbox,
local_default=local_default,
):
return
from rich.console import Console as _Console
_Console(stderr=True).print(
"[yellow]Warning:[/yellow] JS interpreter (`js_eval`) is unavailable "
"under a remote sandbox; it runs in local mode only."
)
def _resolve_rubric_text(rubric: str | None) -> str | None:
"""Resolve the rubric from `--rubric` into one string.View on GitHub (pinned to a1af029e6e)
Solutions
- Force-reinstall the package to restore the bundled manifest
- Regenerate the manifest locally so `interpreter.enable_interpreter` exists
- Fix any test stub or patch of `get_option` that drops the entry
Example fix
// before get_option = lambda name: None # stub missing entry // after get_option = lambda name: real_manifest[name] if name == "interpreter.enable_interpreter" else real_manifest.get(name)
Defensive patterns
Strategy: validation
Validate before calling
from deepagents_code.config_manifest import get_option
assert get_option("interpreter.enable_interpreter") is not None, (
"config manifest out of date; reinstall deepagents-code"
) Try / catch
try:
_warn_if_interpreter_disabled_by_sandbox(args)
except RuntimeError as exc:
logger.warning("skipped interpreter sandbox warning: %s", exc) Prevention
- Keep code and bundled data files version-matched (single package manager)
- Run manifest regeneration after schema edits before launching the CLI
- Cover manifest completeness with a unit test asserting required option keys
When it happens
Trigger: `cli_main` startup (or the warning helper directly) with a manifest missing the `interpreter.enable_interpreter` entry — stale manifest after upgrade, stripped build, or a test stub overriding `get_option`.
Common situations: Same as errors [321]/[322]: version-mismatched installs, vendored distributions, development checkouts with an unregenerated manifest.
Related errors
- manifest option 'interpreter.enable_interpreter' is missing;
- interpreter.enable_interpreter is missing from the config ma
- shell.allow_list is missing from the configuration manifest
- thread display options are missing from the configuration ma
- {what} must be absolute: {path}
AI-assisted analysis of langchain-ai/deepagents@a1af029e6e (2026-08-29).
Data as JSON: /api/errors/426b9f311bd3025b.
Report an issue: GitHub.