README.md

---
license: mit
---
# Table of Contents

* [InterpreterAtomicFlow](#InterpreterAtomicFlow)
  * [InterpreterAtomicFlow](#InterpreterAtomicFlow.InterpreterAtomicFlow)
    * [set\_up\_flow\_state](#InterpreterAtomicFlow.InterpreterAtomicFlow.set_up_flow_state)
    * [run](#InterpreterAtomicFlow.InterpreterAtomicFlow.run)

<a id="InterpreterAtomicFlow"></a>

# InterpreterAtomicFlow

<a id="InterpreterAtomicFlow.InterpreterAtomicFlow"></a>

## InterpreterAtomicFlow Objects

```python
class InterpreterAtomicFlow(AtomicFlow)
```

This flow is used to run the code passed from the caller.
*Input Interface*:
- `code`
- `language`

*Output Interface*:
- `interpreter_output`: output of the code interpreter
- `code_runs`: whether the code runs successfully or not

*Configuration Parameters*:
- None

**Full credits to open-interpreter (https://github.com/KillianLucas/open-interpreter)
for the usage of code interpreters (package `code_interpreters`) and the function truncate_output()**

I'm extracting the code interpreter part from open-interpreter because the litellm version of open-interpreter
is not compatible with that of the current version of aiflows(v.0.1.7).

<a id="InterpreterAtomicFlow.InterpreterAtomicFlow.set_up_flow_state"></a>

#### set\_up\_flow\_state

```python
def set_up_flow_state()
```

class-specific flow state: language and code, which describes the programming language and the code to run.

<a id="InterpreterAtomicFlow.InterpreterAtomicFlow.run"></a>

#### run

```python
def run(input_message: FlowMessage)
```

Run the code interpreter and return the output.

**Arguments**:

- `input_message` (`FlowMessage`): The input message of the flow.