# Code Execution

> Let the model write code, run it in a sandbox, and use the output as the answer — no more trusting the LLM to compute in its head.

- **Category**: Agentic AI
- **Subcategory**: Tool Use & Environment
- **Canonical URL**: https://designpattern.fyi/patterns/code_execution/

---

## Description
**Intent**: Offload deterministic computation to an actual interpreter instead of hoping the model gets the math right.

**Context**: Your agent does arithmetic, data wrangling, parsing, or other deterministic work. LLMs hallucinate on this stuff. You have a sandboxed Python or JS interpreter available.

**Solution**: The agent emits a code block; a controlled sandbox (Python, JS VM, or container) runs it; stdout/stderr/return value flow back. Repeat within a step budget. The CodeAct approach treats code as the primary action language.


## Use Cases
- The task involves calculations, parsing, or transformations that LLMs reliably hallucinate on.
- A controlled sandbox is available and trusted to run model-emitted code.
- Stdout, stderr, and return values can feed back into the agent loop.





## Trade-offs


### Advantages

- Deterministic compute on top of probabilistic intent — the right division of labor.

- Code is auditable and replayable; the same script can be rerun for debugging.




### Considerations & Drawbacks

- Sandbox security is its own serious engineering problem — weak sandboxes mean arbitrary code execution.

- Very flexible action space increases failure modes compared to a curated tool palette.







---
**Reference**: [Original Source](https://www.agentpatternscatalog.org/patterns/code-execution/)