ChatGPT Automatic Code Review in Gitlab MR — AI Product | Success.
ChatGPT Automatic Code Review in Gitlab MR
ChatGPT Automatic Code Review in Gitlab MR — free, ready-to-import n8n workflow (14 nodes, AI and LLMs).
AI ProductReusable license
from $0
Workflow Intelligence
Understand this workflow before you use it.
Full structure
1 to review
Health 87/100
Full workflow structure analyzed
Every step's connections are present in the source file, so the flow, paths and structure below reflect the real workflow.
This workflow starts from Webhook and works with AI text processing and custom code.
Steps
10
Integrations
1
Connections
8
Complexity
Intermediate
Technical health
87/100
Security
1 to review
OpenAI
Overview
**ChatGPT Automatic Code Review in Gitlab MR**
Category: AI and LLMs · 14 nodes
Nodes/integrations used: chainLlm, httpRequest, lmChatOpenAi, splitOut, webhook
How to use: download the JSON below, then in n8n choose Import from File (or open the demo link, copy the JSON, and paste with Ctrl+V into the n8n canvas). Add your own credentials and activate.
Source: enescingoz/awesome-n8n-templates (CC-BY-4.0). Credentials were removed and files modified for redistribution — see the library's ATTRIBUTION_AND_LICENSES.md for full license text.
After you purchase
Every downloadable file is security-scanned before it becomes available to you — files stay locked until that check passes.
Delivery is via short-lived, authenticated download links issued only after your purchase is confirmed — never a public file address.
Arrows are real connections from the workflow file. Dashed lines attach AI components (models, memory, tools) to the step that uses them.
Select a step to see what it is and how it fits in.
Security3 note(s) to review
0 Critical
0 High
1 Medium
2 Low
0 Info
MediumWebhook accepts unauthenticated requests
Anyone who learns the webhook URL can trigger this workflow with arbitrary data.
Confidence: medium · Heuristic
Suggested fix: Enable header or basic auth on the Webhook node, or verify a signature from the sending service.
LowUser-influenced URL path in HTTP request
The request host is fixed, but part of the path/query is built from data that can originate from a public entry point. Path injection is possible; full SSRF is not.
Confidence: high · Heuristic · Uses incoming request data
Suggested fix: Validate the target against an allow-list of hosts before the request, and block private/loopback/link-local ranges on the n8n host.
LowUser-influenced URL path in HTTP request
The request host is fixed, but part of the path/query is built from data that can originate from a public entry point. Path injection is possible; full SSRF is not.
Confidence: high · Heuristic · Uses incoming request data
Suggested fix: Validate the target against an allow-list of hosts before the request, and block private/loopback/link-local ranges on the n8n host.
HealthHow the technical score is made up
Overall 87/100.
Security82/100
−10: Webhook accepts unauthenticated requests
−4: User-influenced URL path in HTTP request
−4: User-influenced URL path in HTTP request
Structural100/100
Data flow100/100
Error handling40/100
−60: No error handling around external calls
Maintainability100/100
Integration risk100/100
Execution complexity100/100
Technical hygiene signals detected statically from the workflow file. Not a guarantee of correctness or fitness for a particular use.
Execution profileWhat one run touches
Starts from: Webhook
HTTP requests
2
AI model calls
1
Custom code steps
2
OpenAI · model set at runtime
Cost
Pricing unavailable
Model usage is shown above; no price estimate is made.
Simulation
Safe simulation available
A mock walk-through with placeholder data. Nothing real is run or contacted.
DocumentationGenerated from the workflow's structure
What it does
This workflow starts from Webhook and works with AI text processing and custom code.
Services involved: OpenAI.
This description is generated from the workflow's node types and wiring. It does not interpret each node's configuration.
Inputs
Data enters through the trigger(s) below.
Some steps read fields from the incoming request body, query or headers.
Webhook
Main flow
The longest path runs through 9 steps:
Webhook
Need Review1 (If)
Get Changes1 (HTTP Request)
Split Out1 (Split Out)
Skip File Change1 (If)
Parse Last Diff Line1 (Code)
Code
Basic LLM Chain1 (LLM Chain)
Post Discussions1 (HTTP Request)
Outputs
No terminal step writes to an external service or responds to a caller.
Requirements
No credentials are attached in this file. You will need your own n8n credentials for each service below.
OpenAI
Setup
Import the JSON file into n8n (Workflows → Import from File).
Create credentials for each service and select them on the matching nodes.
Run it once manually with test data before activating.
Important considerations
No node is configured to retry or continue on failure.
There is no Error Trigger in this workflow.
No error handling around external calls
Generated automatically from the workflow's structure. Verify behaviour in n8n before relying on it.
Supported conversionsConvert to another provider as a new copy
OpenAI Chat Model → Google Gemini Chat ModelApplies to 1 step · about 58% of settings carry over; the rest needs review
A conversion creates a new copy of the workflow for you to review in n8n. The original is never changed.
Automated static analysis (engine 1.1.0). Nothing was executed. Not a guarantee of correctness or safety.
The seller has not attached downloadable files yet.
Uploaded by the seller and security-scanned before any buyer can download them.
