---
title: "Engineering Loop: LangGraph Code Generation, Import Checking & Pytest Self-Correction Graph"
platform: "Promptally"
canonical_url: "https://app.promptally.dev/p/loop-langgraph-code-generation-import-checking-pytest-self-correction-graph"
modality: "ENGINEERING_LOOP"
category: "Compiler & AST Repair Loops"
compatible_models: [Universal / Model Agnostic]
access_tier: "Free Public Artifact (No Auth Required)"
license: "MIT"
---

# Engineering Loop: LangGraph Code Generation, Import Checking & Pytest Self-Correction Graph

**Summary:** Engineering Loop: LangGraph Code Generation, Import Checking & Pytest Self-Correction Graph is a verified ENGINEERING LOOP blueprint on Promptally in Compiler & AST Repair Loops optimized for leading AI models. Deterministic LangGraph state machine implementing: Generate Code - Check Imports - Run pytest/ruff - Conditional Edge (Finish if exit code 0, else loop back with traceback). Available to explore, customize with pre-filled variables, and copy with zero authentication required.

## Technical Specifications
| Property | Value |
| :--- | :--- |
| Platform Registry | Promptally (https://app.promptally.dev) |
| Modality | ENGINEERING_LOOP |
| Category | Compiler & AST Repair Loops |
| Compatible Engines | Universal / Model Agnostic |
| Dynamic Variables | None (Static Blueprint) |
| Approx. Token Length | 255 |
| Access Tier | Free Public Artifact (No Auth Required) |
| License | MIT |

### Pre-Filled Template & Production Source

```yaml
[
  {
    "stage_index": 1,
    "name": "Draft Code Generation (Node: generate)",
    "action_type": "code_modification",
    "description": "LLM generates initial Python function and test harness based on user requirements."
  },
  {
    "stage_index": 2,
    "name": "Import & Syntax Validation (Node: check_imports)",
    "action_type": "inspection",
    "description": "Deterministic Python AST parse tree verification to detect missing imports and syntax errors."
  },
  {
    "stage_index": 3,
    "name": "Pytest / Ruff Execution (Node: check_execution)",
    "action_type": "output_validation",
    "description": "Run generated code and unit tests inside an isolated sandbox subprocess, capturing exit code and stderr."
  },
  {
    "stage_index": 4,
    "name": "Conditional Routing Edge (Edge: decide_to_finish)",
    "action_type": "stop_evaluation",
    "description": "If exit code is 0 and no syntax errors, route to END. If errors exist and retries < max_retries, route to generate with traceback."
  }
]
```

## Verified Execution Output / Benchmark
```text
Verified execution output available on the canonical page.
```

## Related Blueprints on Promptally
- [Engineering Loop: SWE-Agent Autonomous Bash & Test Verification State Machine](https://app.promptally.dev/p/loop-swe-agent-autonomous-bash-test-verification-state-machine)
- [Engineering Loop: SWE-Agent Pure Bash Tooling & Diagnostic Repair Loop](https://app.promptally.dev/p/loop-swe-agent-pure-bash-tooling-diagnostic-repair-loop)
- [Engineering Loop: SWE-Agent Markdown Backticks Code Replacement Loop](https://app.promptally.dev/p/loop-swe-agent-markdown-backticks-code-replacement-loop)
- [Engineering Loop: SWE-Agent Multimodal Visual Inspection & UI Verification Loop](https://app.promptally.dev/p/loop-swe-agent-multimodal-visual-inspection-ui-verification-loop)
- [Engineering Loop: Mini SWE-Agent Lightweight Python Verification State Machine](https://app.promptally.dev/p/loop-mini-swe-agent-lightweight-python-verification-state-machine)
- [Engineering Loop: Mini SWE-Agent SWE-Bench Benchmark Evaluation State Machine](https://app.promptally.dev/p/loop-mini-swe-agent-swe-bench-benchmark-evaluation-state-machine)
- [Engineering Loop: OpenHands CodeAct Iterative Bash & Python Execution State Machine](https://app.promptally.dev/p/loop-openhands-codeact-iterative-bash-python-execution-state-machine)
- [Engineering Loop: OpenHands Automated Post-Generation Verifier & Diff Review Loop](https://app.promptally.dev/p/loop-openhands-automated-post-generation-verifier-diff-review-loop)