Mohammad Saed

304 posts

Mohammad Saed banner
Mohammad Saed

Mohammad Saed

@Moham_Saed

Founder @GateOfAI | Building production-ready agentic workflows & custom LLM pipelines for developers | Sharing practical AI tools & insights

UAE Katılım Temmuz 2022
104 Takip Edilen48 Takipçiler
Mohammad Saed
Mohammad Saed@Moham_Saed·
لو بتبني Agent، أكبر مشكلة واجهتك معاه كانت شو؟ (hallucinations, ينسى المهام, يطلع برا السياق, ولا شي ثاني؟) خلينا نبادل الخبرات.
العربية
0
0
0
2
Mohammad Saed
Mohammad Saed@Moham_Saed·
بنيت SPA Compiler عشان أي مطور (أو حتى غير مطور) يقدر يولد برومبت بهذا الإطار في ثواني، مو ساعات.
العربية
1
0
0
2
Mohammad Saed
Mohammad Saed@Moham_Saed·
معظم الناس بيبنوا AI Agents وبعد أسبوع يرموها لأنها بتخربط وتعمل hallucinations 😩 السر مش في الموديل… السر في System Prompt اللي بيكتبوه. جربت مئات البرومبتات، ولقيت إطار واحد بيفرق النتيجة 180 درجة: S.C.H.E.M.A 🧵
العربية
1
0
0
3
Mohammad Saed
Mohammad Saed@Moham_Saed·
لو مطور وبتبني Agents أو بتشتغل مع LLMs، هالأداة راح توفر عليك عشرات الساعات. جربها وقولي شو عملت 👇
العربية
0
0
0
10
Mohammad Saed
Mohammad Saed@Moham_Saed·
معظم المطورين بيضيعوا ساعات طويلة يكتبوا System Prompts، وبعدين تطلع hallucinations أو ما بتنفع في الإنتاج. تعبت من هالمشكلة، فبنيت SPA Compiler — أداة AI مجانية بتولد برومبتات جاهزة للإنتاج في ثواني باستخدام إطار S.C.H.E.M.A القوي 🧵
Mohammad Saed tweet media
العربية
1
0
0
23
Mohammad Saed
Mohammad Saed@Moham_Saed·
هذا فعلاً من أقوى الميزات اللي صارت في Gemini مؤخراً. القدرة على تنظيم الرسائل + إدارة المهام + الـ Workflows الذكية راح تغير تجربة Gmail كلياً. الخطوة الجاية المهمة هي: إن المستخدم يقدر يبني **Agents** خاصة فيه تعمل على Gmail بشكل متقدم. شو رأيكم، هل بتتوقعوا إن Gemini يصير أداة أساسية للـ Productivity في 2026؟
العربية
0
0
0
49
الجزيرة - تكنولوجيا
بعيدًا عن مزايا «جيميل» المعروفة، يتيح متجر إضافات «غوغل» أدوات تساعد على تخصيص تجربة استخدام البريد وتحسينها من مختلف الأجهزة.. فما أبرزها؟
الجزيرة - تكنولوجيا tweet media
العربية
2
1
25
2.1K
Mohammad Saed
Mohammad Saed@Moham_Saed·
This is exactly why we're building @GateOfAI. Agent-powered tools are no longer "nice to have" they're becoming essential for both developers and educators. I built the **SPA Compiler** specifically to help developers create production-ready system prompts for complex agentic workflows in seconds (supports CrewAI, LangChain, MCP, Claude, etc.). Super excited for AI Skills Fest! Would love to see more sessions on practical agent orchestration. Try the tool: gateofai.com/dev-prompts/ What are you most looking forward to in the fest? 👇
English
0
0
0
10
Microsoft Learn
Microsoft Learn@MicrosoftLearn·
AI is becoming part of how we learn and teach. Join AI Skills Fest (June 8–12) to explore how AI is applied in academic settings, from responsible AI to agent-powered tools and future career paths. Free online event for educators and students. Register early: msft.it/6010vZNas Complete an AI Skills Fest playlist to unlock a free Microsoft Certification exam voucher, earn a Credly badge, and get entered for a chance to win prizes.* *Terms apply: msft.it/6011vZNat
English
4
9
61
3.4K
Mohammad Saed
Mohammad Saed@Moham_Saed·
Already getting good feedback on the SPA Compiler 🔥 If you tried it, tell me: • Which platform did you use it for? (Claude, CrewAI, n8n, etc.) • What did you build? The generations are still free → gateofai.com/dev-prompts/ Repost or Like if you’re a developer working with AI agents 👇
Mohammad Saed@Moham_Saed

Most developers waste hours writing system prompts that still hallucinate or break in production. I fixed that. I built the SPA Compiler, a free AI tool that generates production-ready system prompts in seconds using the S.C.H.E.M.A framework. 🧵

English
0
0
0
58
Mohammad Saed
Mohammad Saed@Moham_Saed·
If you’re a developer building with LLMs or agents, this will save you dozens of hours. Try it and reply with what you built 👇
English
0
0
0
32
Mohammad Saed
Mohammad Saed@Moham_Saed·
Most developers waste hours writing system prompts that still hallucinate or break in production. I fixed that. I built the SPA Compiler, a free AI tool that generates production-ready system prompts in seconds using the S.C.H.E.M.A framework. 🧵
Mohammad Saed tweet media
English
1
0
0
106
Mohammad Saed
Mohammad Saed@Moham_Saed·
ميزة الـ auto-load للـ plugins المحلية من مجلد ⁠.claude/skills⁠ بتسهّل وتسرّع تجربة بناء واختبار الأدوات المخصصة بشكل كبير قبل النشر. تحديث ممتاز للـ DX ومعالجة مشاكل الصور التالفة (Zero-byte) خطوة ذكية لتفادي كراشات مباغتة في الـ CLI.
Claude Code Changelog@ClaudeCodeLog

Claude Code 2.1.157 has been released. 33 CLI changes Highlights: • Plugins placed in .claude/skills auto-load, enabling local plugins to run without marketplace publishing • Sandbox network permission prompts suppressed in auto/bypass for desktop/IDE/SDK to avoid spurious dialogs • Zero-byte or corrupt pasted images no longer crash requests; they become text placeholders to avoid failures Full details are in thread

العربية
0
0
0
74
Mohammad Saed
Mohammad Saed@Moham_Saed·
The sandbox network permission bypass for IDE/SDK is a massive quality-of-life update. Forcing developers to manually approve network requests during complex, multi-step agent runs completely breaks the autonomous workflow chain. Also, auto-loading local plugins via ⁠.claude/skills⁠ without marketplace gatekeeping is exactly how we scale custom internal agent workflows effectively.
English
0
0
0
455
Claude Code Changelog
Claude Code Changelog@ClaudeCodeLog·
Claude Code 2.1.157 has been released. 33 CLI changes Highlights: • Plugins placed in .claude/skills auto-load, enabling local plugins to run without marketplace publishing • Sandbox network permission prompts suppressed in auto/bypass for desktop/IDE/SDK to avoid spurious dialogs • Zero-byte or corrupt pasted images no longer crash requests; they become text placeholders to avoid failures Full details are in thread
English
11
9
253
23.2K
Mohammad Saed
Mohammad Saed@Moham_Saed·
Converting images into clean JSON schemas 100% locally on-device is a massive win for mobile agent architecture. The real bottleneck for pocket agents has always been the compounding latency and context token costs of cloud round-trips for basic tool calling. Moving the core vision-to-structured-data pipeline and skill execution entirely offline change the economics of consumer AI. Edge-native orchestrators are finally becoming viable
English
0
0
0
99
Google Gemma
Google Gemma@googlegemma·
A completely local agent that lives right inside your pocket. 📱 Watch Gemma 4 run 100% locally in the Google AI Edge Gallery app. It converts images into JSON schemas, transcribes audio, and uses agent skills to interact with apps, all entirely offline.
English
48
103
1.1K
64.9K