frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

My 600 Hours with AI Coding Assistants: A Practical Comparison

2•bv_dev•1y ago
After spending over 600 hours using various AI coding assistants over the past 3 months, I wanted to share my experience for those navigating this rapidly evolving landscape.

What I Mean by "Agentic Mode" First, to clarify: by "agentic mode," I'm referring to the assistant's ability to understand project context, reason through multi-step problems, and autonomously make coherent code changes across files without constant hand-holding. True agency means the tool can maintain context across interactions and execute on high-level directions.

The Current Landscape (May 2025) Augment Code - Current go-to tool despite higher costs

Strengths: Maintains context remarkably well across complex refactors; actually understands project structure; can implement feature requests that span multiple files Weaknesses: More expensive than alternatives ($30/month vs $20 for others); occasional hallucinations when venturing outside codebase context Best for: Complex refactoring tasks and implementing features that span multiple files

Windsurf - Slightly edges out Cursor for agentic capabilities

Strengths: Better context retention than Cursor; decent file traversal; good understanding of code relationships Weaknesses: Can get quite stuck in their full agentic mode as it starts editing things. While they have removed their flow credits part, it is still painful to watch it go completely out of context. Best for: Mid-size projects where you need moderate autonomy

Cursor - Popular but underwhelming for true agentic work

Strengths: Good IDE integration; clean interface; works reasonably well for single-file tasks. I like the ability to Cmd+K and insert a bulk of code in the middle. Also, I like the @Docs feature to bring latest documentation for popular libraries.

Weaknesses: Context falls apart in agentic mode; often loses track of previous instructions; requires excessive prompting Best for: Single-file optimizations and modifications, but not complex cross-file tasks

Claude Code - Declining quality since public beta

Strengths: Used to have superior reasoning and contextual understanding 3 months ago Weaknesses: Super expensive (like always), but recent updates have significantly degraded agentic capabilities; now requires much more hand-holding than before as it goes compleltely off base. Best for: Simple tasks that don't require deep contextual understanding Note: Most disappointing decline in quality - was previously much more capable. I spent $500 in Feb-Mar and thought it was worth.

Cline, Roo, and Aider - Conceptually interesting but practically limited

Strengths: Cline has good terminal integration; Roo offers interesting visualization; Aider has straightforward CLI Weaknesses: All three struggle with maintaining context; limited understanding of project structure; frequent need to repeat instructions Best for: Very simple, isolated coding tasks or experiments

Real-world Performance Differences The gap between these tools becomes most apparent when trying to implement complex features. For example, when asked to "add user authentication with email verification to my Express app":

Augment Code: Identified relevant files, added middleware, routes, and email service integration, then explained how the pieces fit together Windsurf/Cursor: Added authentication to single files I pointed at but needed explicit instructions for each additional component Others: Generally required file-by-file guidance with frequent context reminders

Conclusion If budget isn't a concern, Augment Code currently offers the most truly agentic experience, but still has a long way to go. For more budget-conscious developers, Windsurf slightly edges out Cursor for agentic capabilities, though both still require significant guidance for complex tasks.

Comments

SoMomentary•1y ago
I'm always surprised by people sleeping on GitHub Copilot. Is this because people truly don't find any value in it?
bv_dev•1y ago
I have used Github copilot since their beta release in 2023 and I don't find it anywhere near good these days. Automplete was good, but the industry has moved way beyond 2025. Copilot is slightly worse than Cursor which is itself a pretty average tool now. If you use truly agentic code generation, you won't be able to go back to Github Copilot.
SoMomentary•1y ago
You don't consider copilots agent mode to be agentic? I've had some pretty great results with agent mode + mcp to have it check it's own work.

Show HN: StaleBin – share a link, collect anonymous submissions, no accounts

https://stalebin.app/
1•AndBoh•9s ago•0 comments

Show HN: Aitori, see and govern the AI traffic leaving your machine

https://github.com/truefoundry/aitori/
1•deeptishukla22•1m ago•0 comments

America's Big Cities Are Rapidly Losing Kids

https://www.wsj.com/us-news/cities-losing-families-children-a860a9b1
1•littlexsparkee•1m ago•1 comments

UK's iconic Jodrell Bank telescope faces axe in science cuts

https://www.bbc.co.uk/news/articles/c5y6r9rdxygo
2•ttmbs•2m ago•0 comments

Manage Elasticsearch and OpenSearch clusters from one tiny, secure Go binary

https://github.com/hpoznanski/medulla
1•hpoznanski•2m ago•0 comments

Show HN: Browser CAD for stock-tube structures with generated 3D-printed joints

https://app.trusslab.ai/#/frame/triangle-frame@v1?autoplay=1
1•dajanah•2m ago•0 comments

Do Qwen 3.6 27B quantizations break the pelican?

https://quesma.com/blog/qwen-quantization-quality/
1•stared•3m ago•0 comments

Show HN: PumpProof – Scores a hyped stock's dilution risk from its SEC filings

https://pumpproof.com
3•jekjek•4m ago•0 comments

Where B2B Buyers Ask on Reddit: 30 Days of Scan Data

https://cuescout.com/blog/where-b2b-buyers-actually-ask-on-reddit
1•weeklyrunner•4m ago•0 comments

Workers are crossing job boundaries with AI, OpenAI research shows

https://www.axios.com/2026/07/27/openai-chatgpt-work-specialists
2•giuliomagnifico•6m ago•0 comments

There are no particles, there are only fields (2012)

https://arxiv.org/abs/1204.4616
1•surprisetalk•6m ago•0 comments

Why the Cleanup Sprint Always Loses

https://lamis73.substack.com/p/why-the-cleanup-sprint-always-loses
1•Lamis-A•6m ago•0 comments

Paul Andrews, Founder of 'Contact in the Desert'–Woodstock of U.F.O.s–Dies at 77

https://www.nytimes.com/2026/07/26/us/paul-andrews-dead.html
1•bookofjoe•6m ago•1 comments

97 Hours of Consensus

https://skepticalscience.com/nsh/?#
2•Gravityloss•7m ago•0 comments

The State of X402

https://vibesprings.net/state-of-x402
2•Vibeguy900•8m ago•0 comments

Of 1190 AI agent closing statements, one reports failure

https://github.com/kolesnikov-arch/patchward/blob/main/selfreport/RESULTS.md
1•kolesnikov-arch•8m ago•0 comments

Member of Technical Staff

https://torre.ai/post/2W1pn4Yr-qvid-talent-solutions-member-of-technical-staff-product-engineering
2•cnjeri•8m ago•0 comments

AGI Singer: The Song of My AGI Capability and Boot

https://medium.com/@miho999lv/agi-singer-the-song-of-my-agi-capability-and-boot-7f1beec2e8c5
1•miho999lv•9m ago•0 comments

Why open AI models have become a matter of American national security [Italian]

https://www.dday.it/redazione/58170/perche-i-modelli-ia-open-sono-diventati-una-questione-di-sicu...
1•lorenzomx•11m ago•1 comments

The Red Cross plans to extend its protection to cyberspace

https://www.swissinfo.ch/eng/international-geneva/the-red-cross-plans-to-extend-its-protection-to...
1•giuliomagnifico•12m ago•0 comments

Show HN: Panorama Markdown – Markdown dialect and editor for academic papers

https://github.com/panorama-lab/markdown
1•wang_zuo•13m ago•0 comments

Porting Super Hexagon to the Playdate

https://captnemo.in/superhexagon/
1•chmaynard•14m ago•0 comments

Show HN: A udev implementation in Guile Scheme

https://codeberg.org/goodev/goodev
2•mbakke•15m ago•0 comments

Show HN: Wasp, a full-stack web framework, replaces its DSL with TypeScript

https://wasp.sh/blog/2026/06/15/wasp-typescript-spec
1•matijash•16m ago•0 comments

HTTP Message Signatures with curl

https://daniel.haxx.se/blog/2026/07/27/http-message-signatures-with-curl/
3•chmaynard•20m ago•0 comments

Guide to the Kimi DeltaNet Family of linear attention

https://blog.doubleword.ai/you-could-have-come-up-with-kimi-delta-attention
1•kkm•21m ago•0 comments

Young Adults Are Letting AI Do Their Talking for Them–Even in Person

https://www.wsj.com/tech/ai/ai-chatbot-in-person-social-interactions-d1cb6831
3•thm•21m ago•0 comments

Show HN: Mapify Your Repository

https://codemap.gitbiased.com
1•skyfantom•22m ago•0 comments

Meta Got Everything It Wanted in a Secret Louisiana Data Center Deal

https://www.nytimes.com/2026/07/27/technology/meta-data-center-louisiana.html
1•jryio•24m ago•0 comments

Convergence in LLM Quality and Slowdown in LLM Improvement

https://braddelong.substack.com/p/convergence-in-llm-quality-and-slowdown
1•Anon84•29m ago•0 comments