Maxime Prades

7.6K posts

Maxime Prades banner
Maxime Prades

Maxime Prades

@prades_maxime

Product @Cursor, previously Product @meta, @algolia & @zendesk, 🇫🇷 in CA (🥖🧀🍷🛩️)

San Francisco Katılım Nisan 2009
2.9K Takip Edilen2K Takipçiler
Maxime Prades
Maxime Prades@prades_maxime·
@kr0der Have you tried the new « generate rules » feature?
English
1
0
1
25
Anthony Kroeger
Anthony Kroeger@kr0der·
i think everyone should be subscribed to Cursor whether you use Cursor or not for coding, you need a sub to buy Bugbot. in my opinion Bugbot is just insanely valuable and it's worth buying a $20 sub just to access it. it's $40/m for unlimited PR reviews and the accuracy is insane, you can tell that it digs deep into the code rather than just taking the PR at face value. i haven't even bothered looking for a better AI PR review tool because it's too good already. definitely my most valuable non-codegen AI subscription by far
English
25
11
233
19.5K
Maxime Prades retweetledi
Cursor
Cursor@cursor_ai·
Cursor's code review agent can now learn from activity on PRs to self-improve in real time. 78% of issues found are resolved by the time the PR is merged.
Cursor tweet media
English
37
47
855
52.6K
Maxime Prades retweetledi
pedram.md
pedram.md@pdrmnvd·
men in their 40s used to have cool midlife crisis but now they just have agentic workflows
English
158
677
7.6K
440.5K
Maxime Prades
Maxime Prades@prades_maxime·
@pk_iv Amazing. Will try tomorrow. Why ghostty tho?
English
0
0
0
36
Paul Klein IV
Paul Klein IV@pk_iv·
How to set up a new laptop in 2026: 1. Ask Claude to review your current apps and configs and document it in markdown. 2. Install Ghostty and Claude Code on new laptop. 3. Give Claude the markdown file and let it rip It even remembered to disable natural scrolling!!!
Paul Klein IV tweet media
English
44
23
530
36.2K
Maxime Prades retweetledi
Cursor
Cursor@cursor_ai·
We were able to significantly improve the model quality and cost to serve. These quality improvements come from our first continued pretraining run, providing a far stronger base to scale our reinforcement learning.
Cursor tweet media
English
12
27
763
366.6K
Maxime Prades retweetledi
mars
mars@marsBuilds·
this is the Cursor secret sauce a lot of people think that the Cursor app just sends your prompt directly to model providers. instead we do a bunch of cool magic to make the responses better, more context aware, and faster 😎
edwin@edwinarbus

Matt Maher tested frontier models in Cursor v. other harnesses. Cursor boosted model performance by 11% on average: Gemini: 52% → 57% GPT-5.4: 82% → 88% Opus: 77% → 93% His benchmark measures how well models implement a 100-feature PRD. @cursor_ai consistently outperformed.

English
26
9
370
51.9K
Maxime Prades retweetledi
Nock
Nock@0xNock·
Claude: "Go ahead and paste your API key here when you have it" Me: *pastes API key* Claude: "NO you fucking IDIOT"
English
109
171
7.6K
516.6K
Alex Cohen
Alex Cohen@anothercohen·
@kripster Way more limited but Plaid is impossible to get access to as an individual
English
1
0
0
412
Maxime Prades retweetledi
NWS Bay Area 🌉
NWS Bay Area 🌉@NWSBayArea·
🍃Breezy and gusty winds are expected to begin today with gusts of 20-30 mph throughout the region, and 40 mph or above along the coast and at higher elevations. Secure loose items, use caution if driving high profile vehicles, and check back for updates! #CAwx #BayAreaWX
NWS Bay Area 🌉 tweet mediaNWS Bay Area 🌉 tweet mediaNWS Bay Area 🌉 tweet media
English
1
9
25
7.5K
Andrés Matte
Andrés Matte@andresmatte·
this week i'm building from an island in the Carretera Austral, one of the most stunning places i've ever been to get here i had to take: - taxi - plane - taxi again - bus - car - boat 🛶 there's almost no signal out here, but thanks to @Starlink, i can still work
Andrés Matte tweet mediaAndrés Matte tweet mediaAndrés Matte tweet media
English
7
0
58
3K
Andrés Matte
Andrés Matte@andresmatte·
If you have any feedback, I'd love to hear it!
English
2
0
0
315
Andrés Matte
Andrés Matte@andresmatte·
13 product updates in Kapso! 1. You can now use GPT-5.2 and Opus 4.6 in the Kapso Agent. 2. Playbooks: pre-built prompts that you can send directly to the Kapso Agent. 🧵
Andrés Matte tweet media
English
4
1
19
2.4K
Amjad Masad
Amjad Masad@amasad·
Replit creates a separate production (user-facing) database from your development database, and handles safe migrations between them. But sometimes you need to query your production database for data analysis or debugging... So now Agent has read-only access. Enjoy!
Amjad Masad tweet media
English
66
10
280
15.6K
Todd Anderson - AI | SEO
Todd Anderson - AI | SEO@_toddanderson·
I built out Mission Control this weekend for my two companies: - I have 8 agents working for me constantly - Social media scripts are done for Ads, short form vids, and all written - Devs are building autonomously to fix business problems we're encountering - I get updated real-time whenever a task is complete, so I can keep them moving I'll be doing a free webinar answering questions on how I did this and showing you how you can do the same later this week. Like + Comment "Open Claw" for a DM with an invite.
Todd Anderson - AI | SEO tweet media
English
148
27
412
36.7K
Corey Ganim
Corey Ganim@coreyganim·
I taught @openclaw 6 new skills today. Here's exactly what each one does: 1) Brand voice: Writes content that sounds like me, not ChatGPT 2) Proposal generator: Drafts client proposals in deliverable format automatically 3) Stripe invoicing: Creates draft invoices, waits for my approval, then sends 4) YouTube → Tweet: Turns my video transcripts into tactical X posts 5) Costco grocery orders: Reads my list, places the order, done 6) Assessment formatter: Structures our AI Tools Assessment deliverables consistently Total time invested: 1 hour. Time saved per week: 5+ hours. Super high leverage work.
Corey Ganim tweet media
English
13
7
140
13.6K