@ziniman
Account based inIsrael
About this account
- Account based in
- Israel
- Connected via
- Israel App Store
Account-level information from X, not a live location or the device used for a specific post.
Speaker coach for engineers & founders, https://nitter.cf/t.co/ZByPzPz2kJ | ex-AWS Head of DevRel EMEA | Co-founder, Claude Israel UG | Anthropic Claude Community Ambassador
Israel
Joined February 2009
- Tweets5K
- Following510
- Followers2.8K
- Likes2.9K
מדוייק ב - 100%. תקשיבו ל - @urieli17!
Just merged into something special. 🎖️
I am officially a Super Qodoer - part of Qodo's developer ambassador program.
Check it out: qodo.ai/community/super-qodo…
First 5 minutes of PlatforMa conference - “Your Developers Mobed On” (and they are not coming back).
If you don’t get it by now, you are losing touch with the mothership.
@Hilafish1 and @_ediri doing the opening keynote.
אם אתם או חברים שלכם (או יותר טוב, חברות) עובדים / מכירים חבר׳ה מ-AI21 ו/או מטא שהושפעו מהפיטורים האחרונים, אשמח לשמוע מכם. אנחנו בונים משהו גדול ב-AI פה בשרונה ומחפשים מהנדסים סופר חזקים (סיניורים בלבד). בקבוק וויסקי משובח ביותר למי שיפנה מועמד.ת שיתקבל
Boaz Ziniman retweeted
If you missed the last #devrelshow, no need to curl up crying in a hot steamy shower: it was recorded 🎉
Watch @ziniman from SpeakUp discussing all things #devrel & demoing BoothLoader 🙌
youtube.com/watch?v=qqmsiSkZ…
Here's a preview of the show for you!
Looking forward! Thanks for having me on the show.
Join me and my guest @ziniman, CEO & Founder of SpeakUp on May 14 at 9AM ET, while we discuss all things developer relations, and learn more about BoothLoader, which turns any conference booth into an interactive experience with games and live leaderboards, making attendees wanting to stop and engage while turning every event into measurable ROI.
👉 Link in the reply
תפסיקו לבנות אתרים לבני אדם. הלקוח הבא שלכם הוא סוכן AI, והוא לא רואה את הפיקסלים שלכם.
ה-Agentic Web כאן. אם האתר שלכם לא יודע לדבר עם מכונות, אתם פשוט שקופים עבור הסוכנים שיבצעו את הרכישות של מחר.
הפוסט והמדריך החדש שלי על המהפכה השקטה שעוברת על הרשת: ziniman.com/2026/04/making-w…
ai-instruct v2.0.0 is live: 5 web-dev skills for Claude Code in one install.
/plugin marketplace add ziniman/ai-instruct
/plugin install ai-instruct@ai-instruct
Auto-triggered skills for accessibility, SEO/LLMO, Core Web Vitals, AWS deploy, and GA4.
github.com/ziniman/ai-instru…
אל תשתמשו ב - docx skill של Claude Code.
אומר שנית Don’t!
אל תשאלו איך אני יודע… אבל הוא חזיר טוקנים משוגע ויש אלטרנטיבות פשוטות וזולות יותר (תשאלו את Claude הוא ישמח לספר לכם).
מזכיר לי את הימים הראשונים של Lambda. גם הטרמינולוגיה זהה. עוד משהו לשחק איתו.
Now in research preview: routines in Claude Code.
Configure a routine once (a prompt, a repo, and your connectors), and it can run on a schedule, from an API call, or in response to an event.
Routines run on our web infrastructure, so you don't have to keep your laptop open.
ALT Now in research preview: routines in Claude Code
I’m building a new website (details soon) and decided to put my own SEO guide to the test: github.com/ziniman/ai-instru…
The starting point was already solid with a 91 score, but following the steps pushed it to a perfect 100 with almost zero effort. It’s one thing to write the documentation, but seeing it hit those green circles in practice is a different kind of satisfying.
If you're looking to optimize for both search engines and LLMs, the repo is open (make sure you check the README first).
Good take by @itamar_mar on the recent @claudeai code review feature. Looks more like something I could get with a skill than a full review system like @Qodo.
Generating code and verifying code are fundamentally different engineering problems.
Good to see Claude Code recognizing the importance of review.
And does it cost $15-20 per PR ?!@#?
But the real question is:
Should the same system that generates the code also verify it?
In mature systems, we separate concerns:
• Creation systems → generation
• Integrity systems → quality, governance, verification, observability
They operate on different philosophies:
generation optimizes for output, verification for skepticism.
Speed is easy.
Quality is the hard part.
And as code generators evolve, teams will want the freedom to switch between them.
That’s why we believe the winning stack will look like:
Claude + Qodo = speed + quality = velocity
(Claude reviewing Claude risks shared blind spots and even $$ spend?)
אם עוד לא חידשתם את ביטוח הרשות לנזקי מלחמה, תמצאו את האישור משנה שעברה (חפשו בדואר שלכם את ההצהרה), תנו את הקובץ ל - Claude Cowork ותפתחו לו את הטופס למילוי. הוא כבר יעשה את הקסם! יש פה אפילו את הפרומפט.
There's a wall every developer hits when building with AI - the moment you need something outside your expertise, you can't tell if the output is right or just confident.
ai-instruct: drop-in guides that give your AI assistant the domain context it needs to actually get it right.
Works with @claudeai, @cursor_ai, @Copilot, @windsurf, @Base44, @Lovable and more.
Early days - 4 guides so far, contributions from your own experience very welcome.
npx ai-instruct init → github.com/ziniman/ai-instru…