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: Adkumo – Marketing Agent for SaaS

https://www.adkumo.com
1•gustavoauma•2m ago•0 comments

Real-time map of Great Britain's rail network

https://www.map.signalbox.io
1•scrlk•3m ago•0 comments

SW Agent – an outbound-only connector for browser-based PostgreSQL IDEs

2•Vivek-KY•5m ago•0 comments

Bad Epoll (CVE-2026-46242)

https://github.com/J-jaeyoung/bad-epoll
1•birdculture•7m ago•0 comments

Show HN: ScreenCI – Always up-to-date product videos from E2E tests

https://screenci.com/
4•ollipal•15m ago•0 comments

Reliability Issues with Jira and Confluence

https://jira-software.status.atlassian.com
1•amaccuish•15m ago•0 comments

LinguaHover – Hover Any Word, Translate Instantly – Chrome Extension

https://chromewebstore.google.com/detail/linguahover-translator-pr/dnkpfkopeeilpfgnehheceebamlcbcgc
1•titan_tokyo•19m ago•0 comments

Show HN: Mochi Analytics

https://www.mochianalytics.com
1•devtanna•20m ago•0 comments

In the Weights

https://intheweights.com
1•dash2•23m ago•0 comments

Bounding the Blast Radius: A Survey of Prompt-Injection Defenses for LLM Agents

https://fabraix.com/blog/nobody-has-solved-prompt-injection
1•ibrahim_abdu777•24m ago•0 comments

What Happens When AI Agents Can Pay Their Own Bills

https://self-sovereign-agent.github.io/
1•_pdp_•29m ago•2 comments

Higher blood glucose levels linked to faster brain aging

https://medicalxpress.com/news/2026-07-higher-blood-glucose-linked-faster.html
1•XzetaU8•30m ago•0 comments

Supply Chain Attack Campaign PolinRider

https://socket.dev/supply-chain-attacks/polinrider
1•_____k•30m ago•0 comments

Poll: Is there an AI bubble?

1•networked•31m ago•3 comments

Generative AI without guardrail can harm learning:Evidence from high school math

https://www.pnas.org/doi/10.1073/pnas.2422633122
2•teleforce•31m ago•0 comments

Tribe in Colorado brings utility scale solar project online

https://www.npr.org/2026/07/06/nx-s1-5779756/despite-stiff-political-headwinds-tribe-in-colorado-...
2•geox•36m ago•0 comments

Claude Code deletes conversations after 30 days

1•mieubrisse•37m ago•2 comments

Adding Homemade TLS to a Homemade Web Server

https://www.dmytrohuz.com/p/adding-homemade-tls-to-a-homemade
1•dmyhuz•41m ago•0 comments

Euclid scientists discover most distant known quasars

https://www.euclid-ec.org/most-distant-quasars/
1•robin_reala•42m ago•0 comments

Ask HN: What is your go-to prompt to prove AI can be wrong?

1•throwaway_887•45m ago•0 comments

Ask HN: Do you guys struggle with finding a right person for your FDE role?

1•getelementbyiq•52m ago•0 comments

Experts offer advice on performing endurance events in excessive heat

https://news.northeastern.edu/2026/07/02/cycling-in-extreme-heat-tour-de-france/
1•ano-ther•53m ago•0 comments

Read Rust Like a Chinese Character

https://dawaba.pages.dev/writing/sometimes-read-rust-like-a-chinese-character/
1•saint-evan•53m ago•1 comments

A Self-Improving Code World Model Agent

https://jdsemrau.substack.com/p/a-self-improving-code-world-model
1•ph4rsikal•57m ago•0 comments

Show HN: Kiwi – Run agentic dev loops in the cloud, keep keys on your laptop

https://github.com/ibreakthecloud/kiwi
1•harshvkarn•58m ago•0 comments

Plex debuts 5-year membership pass for $250

https://arstechnica.com/gadgets/2026/07/250-used-to-get-you-a-lifetime-plex-pass-now-you-get-a-fi...
1•AndrewDucker•58m ago•0 comments

Australia prosecuting Amazon for unilaterally modifying Prime terms [video]

https://www.youtube.com/watch?v=DV2-GLJ2FnM
1•dataflow•1h ago•1 comments

Show HN: I built notion database based no code form builder

https://ndbforms.myurll.in/
1•nookeshkarri7•1h ago•0 comments

The Hitchhiker's Guide to Agentic AI

https://arxiv.org/abs/2606.24937
2•jonbaer•1h ago•0 comments

Show HN: TV Time to Serializd Importer

https://chromewebstore.google.com/detail/tv-time-→-serializd-impor/cpcfkgcidilfibakiaohbekkanlk...
1•Gooblebrai•1h ago•0 comments