https://xitter.twitterapi.xyz/t.co/46T9Mk2rnp
Bangalore
Joined June 2011
- Tweets10.2K
- Following762
- Followers169
- Likes30.9K
Naveen retweeted
Free Stanford Course: "CS193T Thinking with AI"
No prior experience or technical knowledge required
All course materials including slides and weekly hands-on assignments
web.stanford.edu/class/archi…
Naveen retweeted
If you want to learn how AI agents work, I highly recommend Harvard’s CS2680.
It covers how Claude Code works, how to build your own agent, and how to serve LLMs on GPUs.
cs2680.seas.harvard.edu/
Naveen retweeted
AI agent–enabled coding is quietly becoming the new SDLC.
Software development just had its biggest shift since the GUI.
Planning.
Coding.
Testing.
Deployment.
Agents are starting to handle all of it.
Here’s the shift most engineers haven’t noticed yet 👇
Old model: SDLC
• sequential phases
• human-driven execution
• testing happens after development
• changing requirements break timelines
Everything moves step → by → step.
New model: ADLC (Agent-Driven Lifecycle)
• agents write, refactor, and test code
• multiple tasks run in parallel
• requirements evolve dynamically
• feedback loops happen in real time
Instead of a pipeline…
You get a live development system.
6 major shifts happening right now
1️⃣ Driver
Human execution → Autonomous agents
2️⃣ Planning
Fixed scope → Evolving goals & PRDs
3️⃣ Development Speed
Sequential handoffs → Parallel sub-agents
4️⃣ Testing
Post-development QA → Continuous testing
5️⃣ Adaptability
Mid-cycle chaos → Real-time re-planning
6️⃣ Feedback Loop
End-of-project retros → Live monitoring
Some early signals are already here.
According to agentic coding reports:
• teams at Wiz and CRED doubled execution speed
• large-scale repos are being modified autonomously
• complex implementations completed in hours instead of days
How engineers should adapt
1️⃣ Start with one agent
Automate testing first.
2️⃣ Learn to write clear PRDs
Agents execute exactly what you define.
3️⃣ Introduce parallel sub-agents
Break one large task into smaller workstreams.
4️⃣ Review outcomes, not every line of code
5️⃣ Build live feedback loops
Agents should detect issues before you do.
The future of software development isn’t just faster coding.
It’s agent-driven systems building software.
#AI #AIAgents #SoftwareEngineering #SDLC #GenAI #AIEngineering
Naveen retweeted
Highly recommended! 🚀 Stanford's CS329Z: Engineering AI Agents (Fall 2026), now open and led by Diyi Yang (@Diyi_Yang), is a great resource for anyone interested in AI agents.
One of the most interesting takeaways is the history and challenges of AI agents, as well as the emerging topic of proactive agents.
youtube.com/watch?v=DGI4vucR…
Naveen retweeted
"Microsoft's Mark Russinovich just changed the AI game..."
Naveen retweeted
If you're building LLM apps, they'll need more than basic monitoring.
In this tutorial, Jessica teaches you how to build end-to-end LLM observability in FastAPI with OpenTelemetry.
You'll trace retrieval, prompt building, model calls, token usage, cost, and evaluation signals.
freecodecamp.org/news/build-…
Naveen retweeted
Long-horizon agents can automate tasks that take days or weeks, but much can go wrong. Kiran Prakash explains how durable execution allows agents to recover.
martinfowler.com/articles/du…
Naveen retweeted
The Bitter Lesson of Agent Memory
arxiv.org/abs/2610.29823
⤷ tldr: Invest in better search tools for agents, not necessarily more sophisticated memory architectures.
Naveen retweeted
Eval-driven development (EDD) is TDD for AI agents. Great article from @LangChain
Stop tweaking prompts and guessing whether things got better. Measure every change against a fixed set of evals.
Here’s how to do it 🧵Read the thread to learn.
#agenticai #AIAgents
langchain.com/blog/what-is-e…
Naveen retweeted
🚨ANTHROPIC ACABA DE PUBLICAR UNA GUÍA DE SEGURIDAD DE 36 PÁGINAS Y EL MENSAJE ES CLARO: DEJA DE CONFIAR EN TUS AGENTES DE IA.
Si trabajas con Claude Code, servidores MCP o cualquier herramienta de automatización, sigue leyendo.
El tiempo que tenías para reaccionar ante un ataque ya casi no existe.
Con IA, el salto de una vulnerabilidad a un exploit funcional ya no tarda meses, tarda horas. Y cuesta apenas unos dólares.
Los agentes además abren una superficie de ataque completamente nueva: desde herramientas envenenadas hasta manipulación directa de su memoria de contexto.
El concepto más valioso de toda la guía es esta pregunta:
Este control hace el ataque imposible, o simplemente más lento?
Un atacante automatizado no se cansa.
Los límites de tasa y el 2FA son obstáculos, no muros. Si quieres seguridad real a la velocidad de la IA, necesitas barreras duras y defensas que también sean automáticas.
Lo que Anthropic recomienda para proteger tus agentes:
→ Considera tus API keys estáticas como ya comprometidas. Usa tokens que expiren en minutos.
→ Limita explícitamente lo que cada herramienta puede hacer, no lo que puede ver.
→ Ejecuta en entornos aislados cualquier agente que procese emails, webs o datos externos.
→ Los permisos deben asignarse por tarea, no de forma permanente.
Enlace a la guía en el 🧵↓
🚨 A FREE AI HACKER JUST WENT OPEN SOURCE, AND IT ATTACKS YOUR APP EVERY TIME YOU SHIP A RELEASE
Over 49,000 stars on GitHub and still climbing.
It doesn't hand you a list of possible flaws. It shows you exactly where someone could get in.
It's called Shannon, an open source AI pentester.
You point it at your app and it starts attacking it the same way a real attacker would.
What it does:
→ Attacks with working exploits, not theories
→ If it can't exploit a flaw, it leaves it out of the report
→ Plugs into GitHub Actions or GitLab CI and runs on its own every release
→ Generates a PDF report ready to share with your team
→ Exports in SARIF so flaws show up inside your dev tools
→ Works with Claude, GPT, Grok, DeepSeek or a local model
This week PHP2Shell and Node2Shell dropped, flaws that let anyone run code on a server without a password.
Patching covers what's already known.
Shannon shows you what nobody has reported yet.
Not long ago a pentest was a yearly event.
Now your app gets one every time you update it.
Open source. Free.
Repo in the first comment 👇
Naveen retweeted
Building an Agentic Knowledge Graph for Billions of Events (Members only)
levelup.gitconnected.com/bui…
#MachineLearning #ChatGPT
Naveen retweeted
I'm sharing with you one of the best papers I've read on Harness Engineering
it takes apart Claude Code (Opus 5.5), Codex, Gemini CLI and more, and boils all of them down to one line:
agent = model + harness
what you take away:
> what a harness is, in plain terms
> how modern coding agents are built under the hood
> the patterns they all share
> the direction the whole field is moving
> a checklist for building your own
the model collects the applause
the harness decides whether the agent actually works
loop and graph engineering come next, and that part lives in my article below
Naveen retweeted
Deep learning terms can be a bit confusing when you're first learning them.
In this article, @Roland_Sankara explains CNNs, RNNs, and Transformers with beginner-friendly mental models.
You'll learn how neural networks work, why CNNs focus on spatial patterns, how Keras helps you build models, and lots more.
freecodecamp.org/news/cnns-r…
Naveen retweeted
Google’s new open-source plugin for AI coding agents.
"Design with an elephant and review with a goldfish"
GitHub Repo: github.com/google/elephant-g…
Naveen retweeted
Google just released Elephant-Goldfish Model (EGM)... an opensource plugin with an interesting approach to design reviews
- An elephant agent designs with full project context
- Goldfish agents review with fresh context, challenging assumptions and finding blind spots
All before writing a single line of code
I really like the idea of having a different agent review the design instead of letting the same one approve its own work
Works with
- Claude Code
- Codex
- Antigravity
github.com/google/elephant-g…
Naveen retweeted
AI agents typically only do something after you ask them to, which often works just fine.
But here, Darsh shows how to build a local AI assistant that runs on a schedule.
You'll use Python, Ollama, Qwen, LangChain, cron, and Task Scheduler to run daily agents for common recurring tasks.
freecodecamp.org/news/how-to…
Naveen retweeted
An MCP server exposes tools, resources, and prompts to an AI app using a standard protocol.
And here, @a0m0rajab shows you how to build one and connect it to ChatGPT.
You'll create a todo app with Node.js, the MCP SDK, Supabase Auth, a database, OAuth flow, DigitalOcean deployment, and more.
freecodecamp.org/news/how-to…
Naveen retweeted
Thousands of Google engineers use this open-source plugin to draft specs with deep-context agents and stress-test them with zero-context subagents before writing code.
egm.dev/
Naveen retweeted
How to transform your software development lifecycle with AI
claude.com/resources/article…
