iflytek/astron-agent · error · CustomException

CODE_NODE_RESULT_TYPE_ERROR

CODE_NODE_RESULT_TYPE_ERROR

Error message

Code return type is not str, please check the type!

What it means

After the sandboxed code runs, _check_and_set_variable_pool validates each declared output variable against the node's output schema. When an output variable is declared with type "string" but the returned dict holds a non-str value (int, list, None, etc.), it raises CODE_NODE_RESULT_TYPE_ERROR. This catches contract drift between what the code node's output schema promises and what the script actually returns.

Solutions

1. Open the code node and coerce the offending variable to str before returning: `return {"result": str(value)}`.
2. Or change the output variable type in the node schema to match the actual runtime type (e.g. integer).
3. Add explicit checks in the script for None/error paths so the variable is always a string.
4. Re-run the code node in debug mode to inspect the raw returned dict and confirm each declared type.
5. Note: bool is a subclass of int but not of str — ensure string outputs aren't accidentally booleans.

Example fix

// before
def main() -> dict:
    return {"name": 123}

// after
def main() -> dict:
    return {"name": str(123)}

outputs = {"name": value}
for k, t in declared_schema.items():
    if t == "string" and not isinstance(outputs.get(k), str):
        raise TypeError(f"output '{k}' must be str, got {type(outputs[k]).__name__}")

def is_str_output(v) -> bool:
    return isinstance(v, str)

try:
    result = code_node.async_execute(...)
except CustomException as e:
    if e.err_code == CodeEnum.CODE_NODE_RESULT_TYPE_ERROR:
        log.error("code node output schema mismatch: coerce output to declared type")
        raise

When it happens

Trigger: The code's main() returns a dict whose value for an output variable declared as "string" is not a Python str — e.g. returns 42 for a field typed string, or None when a branch fails to set the value.

Common situations: User edits the output schema to string but forgets to str() the value in code; JSON-parsed numbers passed through as-is; None returned on an error path; language coercion differences inside the sandbox (e.g. numbers for IDs like zip codes).

Understand the failure class

Background: Type mismatch errors: IllegalArgumentException, TypeError and type guards across 150 open-source libraries — this error's family across 150 libraries.

Related errors

• 21602
• WORKFLOW_VERSION_PUBLISH_FAILED
• 8008
• 8008
• WORKFLOW_PROTOCOL_NODE_INFO_CANNOT_EMPTY

AI-assisted analysis of iflytek/astron-agent@5e758547a8 (2026-09-12). Data as JSON: /api/errors/fc6ac8f455d527b1. Report an issue: GitHub.

Appendix: source

Thrown at core/workflow/engine/nodes/code/code_node.py:229

        outputs = {}
        for var_name in self.output_identifier:
            var_type = variable_pool.get_output_schema(
                node_id=self.node_id, key_name=var_name
            ).get("type")

            # If variable not in result, use existing value from variable pool
            if var_name not in code_result_dict:
                final_result = variable_pool.get_variable(
                    node_id=self.node_id, key_name=var_name, span=span
                )
                outputs.update({var_name: final_result})
                continue

            # Type validation based on expected output schema
            match var_type:
                case "string":
                    if isinstance(code_result_dict[var_name], str) is False:
                        raise CustomException(
                            CodeEnum.CODE_NODE_RESULT_TYPE_ERROR,
                            "Code return type is not str, please check the type!",
                        )
                case "integer":
                    if isinstance(code_result_dict[var_name], int) is False:
                        raise CustomException(
                            CodeEnum.CODE_NODE_RESULT_TYPE_ERROR,
                            "Code return type is not integer, please check the type!",
                        )

                case "number":
                    if isinstance(code_result_dict[var_name], (int, float)) is False:
                        raise CustomException(
                            CodeEnum.CODE_NODE_RESULT_TYPE_ERROR,
                            "Code return type is not integer or float, please check the type!",
                        )

                case "boolean":

View on GitHub (pinned to 5e758547a8)