Jakia A.

When you take "Actual dream don't let you sleep" seriously 😳 103° fever and still up🤒

slowly accepting the fact that life doesn't always go the way we want

Claude Code can save you hours. Or waste them if you do not get the mental model. The problem is not syntax. It is the execution model. These are every Claude Code concept out there: 1. What Claude Code is It does not just answer questions. It can take actions on your machine. 2. The terminal That black screen is simply where Claude Code runs. 3. Prompts You tell it what to do in normal language. 4. Permissions Claude can change files, run commands, and edit code, so permission control matters. 5. Tools Read, write, and bash are the core tools behind what it can do. 6. Context window This is the working memory for the current session. 7. Conversation history / resume You can return to earlier sessions instead of starting from zero. 8. Token usage / cost tracking Usage has a cost, so it helps to know what is consuming tokens. 9. CLAUDE.md This is where you define project rules, preferences, and conventions. 10. Memory usage Claude can remember some patterns and preferences across sessions. 11. Compact context When context gets too large, Claude can compress it and keep the important parts. 12. Models Different models trade off speed, cost, and capability. 13. Denying access to files Some files should stay out of bounds completely. 14. Flags Startup options let you control how Claude behaves in a session. 15. Extended thinking For harder tasks, Claude can spend more time reasoning before acting. 16. Slash commands These are shortcuts for common actions inside Claude Code. 17. Skills Skills give Claude better instructions for specific kinds of work. 18. Hooks Hooks run scripts automatically at key moments without using AI tokens. 19. MCP servers MCP connects Claude to external tools, not just local files. 20. Sub-agents Claude can hand off focused tasks to separate agents with separate context. 21. Agent teams For larger builds, multiple agents can collaborate together. 22. Image / screenshot support Sometimes pasting a screenshot is faster than describing the issue. 23. Checkpoints / undo If Claude goes the wrong direction, you can roll back. 24. Git integration You need version control if Claude is making real changes. 25. Headless mode / CLI mode Claude can run more independently without constant interaction. 26. Claude Max vs API How you pay affects how you use it. 27. Worktrees Separate workspaces let Claude handle parallel tasks without collisions. The hard part of Claude Code is not the commands. It is understanding what it can access, what should stay in context, what should be excluded, and when work should move into a new session, sub-agent, or worktree. Which concept took you the longest to understand? youtu.be/ZlDnsf_DOzg?is… --- ⫸ꆛ Join 46,000+ engineers and get instant access: - Free 30-min Zero to Hero AI Agents Training + 50 Best Practices Guide (56 pages) + How to Build Production AI Agents Guide (32 pages) Grab it here: maryammiradi.com/free-ai-agents…

RAG isn't a solution. It's a starting point. Naive RAG breaks on relational questions. Graph RAG breaks on dynamic data. Agentic RAG breaks your budget if you use it wrong. Wrote about when to use each one → tinyurl.com/45whyaea

Dear @X algo, please #connect me with ppl interested in: - Software Engineering 💻 - Frontend 🎨 - Backend ⚙️ - Freelancing 💼 - Web development 🌐 - System Architecture ⚒️ Let's connect 🫶

@insharifain @grok please tell me what this post indicates. When is the first Ramadan in KSA.

TOMORROW: The Investigation of the Ramadan 1447/2026 Crescent اللهم بلغنا رمضان

Civic duty done ✅ Participated in democracy. May our nation move forward with peace, unity, and progress. Wishing for a brighter, stronger future for everyone ✨ #BangladeshElections2026

The most expensive way to solve a problem is with AI

@JobFound5 interested

Role: Interns Stipend: $28 - $31 per hour Location: Remote Let us know if you are Interested 👇

🛠️📎 60 Lessons on 𝗔𝗜 𝗔𝗴𝗲𝗻𝘁𝘀 — In One 𝗖𝗵𝗲𝗮𝘁𝘀𝗵𝗲𝗲𝘁. Everything you need before you start 𝗯𝘂𝗶𝗹𝗱𝗶𝗻𝗴: frameworks, issues, fixes, and examples ⬇️ 》 What are AI Agents? ✸ Software systems powered by LLMs ✸ Handle tasks with autonomy, decision-making & adaptability → Think: Researcher, Planner, Code Assistant, Fraud Detector 》 Types of AI Agents ✸ Fixed Automation ✸ ReAct & RAG ✸ Tool-Enhanced ✸ Memory-Enhanced ✸ Self-Reflecting & Self-Learning → Pick the right type based on your task’s complexity & volatility 》 10 Questions Before You Build One ✸ Is the task frequent? ✸ Does it need adaptability, memory, or real-time knowledge? ✸ What’s the cost-benefit ratio? → Don't build agents where a script will do. 》 Real-World Use Cases ✸ Clinical History Search Engine ✸ Protocol Summarizer ✸ Predictive Maintenance ✸ SOAP Notes Generator ✸ Inventory Management AI → These aren't demos — they're deployed. 》 Top Frameworks (Python) ✸ LangChain • LangGraph • CrewAI ✸ OpenAI Swarm • Pydantic AI ✸ AutoGen • Vertex AI Agent Builder → All work great with LLMs like GPT-4, Claude, and LLaMA 3 》 How to Measure Agent Performance ✸ Speed – Latency per tool call ✸ Cost – Token use, API frequency ✸ Success – Task & output format accuracy ✸ Tool Use – Selection & argument success → Metrics make the agent real, not just impressive 📌 Common Pitfalls & Fixes ✸ Dev Issues → Poor prompts, weak evaluation ✸ LLM Issues → Planning fails, tool calling errors ✸ Production Issues → No guardrails, scaling & infinite loops → Each has a fix. Each costs you if ignored. 🧾 I turned this into a full visual cheat sheet. github.com/kimtth/softwar… ≣≣≣≣≣≣≣≣≣≣≣≣≣≣≣≣≣≣≣≣≣≣≣≣≣≣ ⫸ꆛ Want to build Real-World AI Agents? Join My 𝗛𝗮𝗻𝗱𝘀-𝗼𝗻 𝗔𝗜 𝗔𝗴𝗲𝗻𝘁 𝟱-𝗶𝗻-𝟭 𝗧𝗿𝗮𝗶𝗻𝗶𝗻𝗴 trusted by 1,000+ builders worldwide! ✔ Only basic Python required. ➠ Build Agents for Healthcare, Finance, Smart Cities & More ➠ Master 5 Modules: 𝗠𝗖𝗣 · LangGraph · PydanticAI · CrewAI · OpenAI Swarm ➠ Includes 9 Real-World Projects 👉 𝗘𝗻𝗿𝗼𝗹𝗹 𝗡𝗢𝗪 (𝟱𝟲% 𝗢𝗙𝗙): maryammiradi.com/ai-agents-mast…

Not all strength looks the same. Sometimes it’s just staying quiet and holding on. Sometimes it’s getting back up when you’re tired of everything. People see the fire in your eyes, but they don’t know what you went through to keep it burning. I don’t always feel strong. I don’t have it all together. But I keep going. Because deep down, I don’t want to give up on myself.

Most devs learn to call APIs. Fewer learn to respect them. REST isn't just about endpoints — it's about clarity, intent, and treating every request like a conversation. Understanding verbs, status codes, and CORS isn't extra. It's the job. #buildinpublic #WebDev #APIDesign

You don’t need loud victories to prove you’re moving. Some days, just showing up is the bravest thing you do. You finish your tea. You write one more line. You carry on. Not because the world cheers you on — but because you told yourself you would. If that’s you, you’re not alone. — Somewhere between logic and longing, May 30, 2025

The only real failure is abandoning your principles.

@PhilOfLife_ Couldn't agree more! Both in personal and professional life

Some people are only loyal to you in front of you.

Coding, coffee breaks, and the occasional samosa. Serene evenings, punctuated by the quiet architecture of the city. Some evenings feel like an interlude between significant challenges. 😇🌿

@UxDamini_ The lighter version is lit 🔥

Light or dark? Which one will you pick ?

@ProfTomYeh Looking forward to the livestream. Already signed up 🙌

Can you draw 𝗠𝗖𝗣 𝘃𝘀 𝗙𝘂𝗻𝗰𝘁𝗶𝗼𝗻 𝗖𝗮𝗹𝗹𝗶𝗻𝗴 by hand ✍️? Join our livestream next Wednesday (5/20) lu.ma/tjpnlar6

@OpenAI Need go purchase pro plan for this!

Introducing a research preview of Codex in ChatGPT openai.com/live

@sama Looking forward to the new random name that top the charts!

we will name it better than chatgpt this time in case it takes off

soon we have another low-key research preview to share with you all