GitHub Copilot Reviews & Product Details

I like how easy it is to just provide a prompt and have the work done in seconds. When needed, it can take my input and make changes directly in my file, instead of me having to go back and forth copying and pasting specific pieces of code. Review collected by and hosted on G2.com.

At times your intended task may not be done correctly until you look deep into the logic and find out that what it was supposed to do it didn't do instead it found a simpler hard coded way out. Review collected by and hosted on G2.com.

What I like most about GitHub Copilot is that it helps me stay focused while coding. Instead of spending time looking up syntax or writing repetitive code, I can concentrate on solving the actual problem. It's especially useful for generating boilerplate code, suggesting functions, and helping me learn new libraries or frameworks more quickly. I always review and understand the suggestions before using them, but overall it saves time and makes development much more efficient. Review collected by and hosted on G2.com.

One thing I dislike about GitHub Copilot is that its suggestions are not always accurate or optimized, especially for complex logic or unfamiliar codebases. Sometimes it confidently generates code that looks correct but contains subtle bugs or doesn't follow the project's coding style. It can also make it tempting to accept suggestions without fully understanding them, so I think it's important to review and test everything before using it. Overall, it's a great productivity tool, but it still requires human judgment. Review collected by and hosted on G2.com.

I use GitHub Copilot on a daily base, it imporves my productivity by generating code suggestions when I work with VS code, and provide useful solutions for creating a frameworks. Saved me from repetitive coding tasks especailly recently on a work with building Data quality framework with Python. Review collected by and hosted on G2.com.

Not always provide the answers correctly and directly, need to review carefully and test the code using different data. Review collected by and hosted on G2.com.

What I love most about GitHub Copilot is how much it speeds up my daily workflow. It’s incredibly good at predicting repetitive boilerplate code and standard function structures. Instead of constantly looking up minor syntax details or typing out tedious setups, I can just write a comment or start typing a function name, and Copilot fills in the rest. It keeps me in the flow state much longer. Review collected by and hosted on G2.com.

Sometimes it generates code that looks perfectly correct at first glance but contains subtle bugs, deprecated syntax, or references to non-existent library methods. If you aren't paying close attention, it's easy to accept a suggestion that breaks the build, meaning you still have to spend time carefully reviewing and debugging its output. Review collected by and hosted on G2.com.

Contextual Autocomplete: It suggests entire blocks of code, functions, and tests by analyzing your current file and open tabs.

Boilerplate Reduction: It handles repetitive tasks like writing unit tests, regex, or standard API calls, allowing you to focus on logic.

Natural Language to Code: You can write a comment describing what you want (e.g., // function to validate email using regex), and it will generate the implementation.

Multi-language Support: It works across dozens of languages including Python, JavaScript, TypeScript, Ruby, Go, and Java.

IDE Integration: It lives directly inside popular editors like VS Code, JetBrains, and Neovim, so there is no need to switch windows. Review collected by and hosted on G2.com.

1. Inaccuracy and "Hallucinations"

Code Quality: It often suggests code that is inefficient, outdated, or uses libraries that don't actually exist (hallucinations).

Bugs: It can generate syntactically correct code that contains subtle logical errors, requiring you to spend more time debugging than if you had written it yourself.

2. Context Limitations

Large Projects: It sometimes "forgets" logic established earlier in a file or fails to understand the broader architecture of a complex project.

Proprietary Logic: It struggles with custom frameworks or internal business logic that wasn't part of its public training data.

3. Privacy and Security

Data Training: Many users are concerned about their code being sent to GitHub's servers to train future models. As of early 2026, some users have expressed frustration over "automatic opt-in" policies for data collection.

Vulnerabilities: There is a risk that the AI might suggest patterns that include known security vulnerabilities (like SQL injection) if they were prevalent in its training set. Review collected by and hosted on G2.com.

Copilot fits seamlessly into VS Code with fast, reliable suggestions that keep my flow uninterrupted, even on larger tasks. It saves time on repetitive work, making it worth the cost. Setup is quick, and getting started feels effortless with minimal learning curve.

I also like that it gives access to multiple AI models. Review collected by and hosted on G2.com.

One downside of GitHub Copilot is that it sometimes feels a bit slower as compared to Cursor, especially when working on larger or multi-file changes.

But Copilot is much cheaper (around $10/month vs $20 for Cursor), and for day-to-day coding, it still covers most needs really well. Review collected by and hosted on G2.com.

What I like best about GitHub Copilot is how it significantly speeds up development by providing real-time code suggestions and auto-completions. It helps reduce repetitive coding tasks, generates boilerplate code quickly, and even suggests entire functions based on context. This allows me to focus more on problem-solving and logic rather than syntax. I also appreciate how it adapts to my coding style over time and supports multiple languages, making it a valuable productivity booster in day-to-day development. Review collected by and hosted on G2.com.

One downside of GitHub Copilot is that its suggestions are not always accurate or optimal, especially for complex logic or edge cases, so you still need to carefully review and validate the code. It can sometimes generate outdated or inefficient patterns, which may lead to technical debt if used blindly. Another limitation is that it may not fully understand the broader context of a large application, resulting in irrelevant or incomplete suggestions. Additionally, over-reliance on it can reduce hands-on problem-solving and learning for developers, particularly for those still building strong fundamentals. Review collected by and hosted on G2.com.

It speeds up development with intelligent autocomplete and code generation. By analyzing your current file and open tabs and using that context in its prompts, it can better help understand the relevant logic and provide more accurate code completions wrt to that Review collected by and hosted on G2.com.

Because it’s model-specific, it ends up consuming more tokens, which limits the overall user experience. Want something solid about token control to optimize the token usage accordingly. Review collected by and hosted on G2.com.

The agent mode in VS Code has fundamentally changed how I handle my day-to-day work supporting complex enterprise platform issues. I use it constantly for:

- Debugging payments & billing issues - When investigating payment failures, invoice discrepancies, or billing escalations, Copilot helps me trace through transaction flows, search for related Jira tickets from similar past incidents, draft investigation summaries, and identify patterns across cases. It keeps track of the troubleshooting steps I've already tried so I don't go in circles.

- Triaging LBP and platform escalations - Copilot helps me analyze repro steps, determine correct routing (e.g., distinguishing Online Jobs issues from LBP based on user-facing workflows vs. enterprise identifiers), and document findings. It retains the routing logic I've taught it so I don't re-explain our criteria every time.

- Jira ticket management - I create, update, search, and link Jira issues directly from my editor. Instead of context-switching to the browser and filling out 10 fields, I describe what I need in plain English and Copilot handles creation with proper categorization, adds comments, links related issues, and transitions statuses.

- Knowledge base maintenance - I use it to search and update our Confluence documentation - drafting troubleshooting guides from raw case notes, keeping routing rules current, and structuring KB articles so the whole team benefits.

- D365 case categorization - Copilot helps me remember and consistently apply complex routing rules (which Reason/Reason Details map to which team, how page type conversions route, etc.) across hundreds of cases.

The MCP integrations mean I have a single pane of glass for code, tickets, docs, Slack, and internal tooling - all without leaving VS Code. Review collected by and hosted on G2.com.

Occasionally it can be overly cautious - asking for confirmation on things I'd prefer it just did. Context window limits mean that on very large debugging sessions with extensive logs or long payment transaction traces, you sometimes need to re-orient it. The inline suggestions in non-agent mode can be hit-or-miss with internal/proprietary systems, though agent mode solves this by actively gathering context first. Review collected by and hosted on G2.com.

Inline suggestions that never break my flow and make it feel completely natural.

I'll start typing a function name and Copilot fills in the entire body, often exactly what I had in my mind. Also, the chat feature (Copilot Chat) is very underrated. When I need to debug a module I wrote three weeks ago and barely remember, asking Copilot "what is this function doing and where might it break?" gives a genuinely useful answer, better than Googling the error message. Review collected by and hosted on G2.com.

It occasionally gets overconfident with framework-specific code. I've had it suggest deprecated Next.js patterns or generate Mongoose query syntax that was slightly off. Always needs a review pass , you can't accept suggestions blindly.

Also, on large monorepo projects, it sometimes loses context and starts suggesting code that contradicts what's defined two files away. Better cross-file awareness would make it 10/10 tool. Review collected by and hosted on G2.com.