I've always been the gamer in the family. When my son was born nearly 11 years ago, gaming was one of the things I looked forward to sharing with him. Pulling up a chair next to me, he would watch as I played Sea of Thieves with friends, often wearing the too-big-for-his-head headset to talk […]

I built a voice assistant that controls 45 AI tools. I say "Hey Jarvis" and it executes. What It Does Command Action "generate content" Creates YouTube scripts for 9 channels "research quantum computing" Deep research via Tavily + AI "write email about meeting" Drafts email, copies to clipboard "start focus" Starts Pomodoro + blocks apps "code review" Reviews git diff with AI "summarize" Summarizes clipboard content "find file tax PDF" Natural language file search Architecture Mic → Whisper (offline) → Intent Classify → Router → Ollama → say (TTS) Key Features Offline speech (Whisper local) Wake word: "Hey Jarvis" Global hotkey: Ctrl+Space Command chaining: "research AI then write blog" Memory across conversations Hindi + English Setup brew install portaudio pip install SpeechRecognition pyaudio openai-whisper python voice_commander_pro.py 🔗 github.com/amrendramishra/ai-tools 🌐 amrendranmishra.dev

hi, i am Aniruddha Adak, a full-stack developer from kolkata who spends way too much time building things with ai tools, shipping apps, and reading way too many github readmes at 2 am. i built 27 apps in 45 days using no-code and ai tools last year. that experience taught me one thing very clearly: the landscape of ai tooling for developers is moving insanely fast, and it is genuinely hard to keep up. so i sat down and did something about it. this is my deep research post on every ai tool, agent, builder, reviewer, and framework that developers, software engineers, and ai engineers should actually know about right now. i have organized it into categories so you can find what you need quickly. no fluff. just the tools, their sites, and what they do. why i wrote this i keep seeing developers waste time because they do not know the right tool exists. someone is manually reviewing pull requests for a week straight, not knowing coderabbit exists. someone else is hand-writing supabase schemas when emergent can do it in seconds. another person is spending days on a landing page when v0 can scaffold it in one prompt. this post is my attempt to fix that. i went through github repositories, dev communities, product hunt launches, and research aggregators to compile this. it is long. that is intentional. bookmark it. section 1: ai-native ides these are not just editors with a chatbot plugged in. these are environments built from the ground up around how language models think and work. tool site what it does cursor https://www.cursor.com forked vscode, codebase-aware context windows, multi-file edits with copilot-style background indexing windsurf https://windsurf.com cascade ai agent that writes files, runs terminal checks, and fixes things in real-time zed https://zed.dev built in rust with gpui, super low latency, native multiplayer coding support replit https://replit.com cloud ide with a full autonomous agent that runs inside serverless virtual workspaces google antigravit

Hey everyone, I change servers more often than I probably should. A discounted VPS or a good coupon is usually enough to convince me, but manually recreating the same web stack every time stopped being fun a long time ago. That is why I built HostShift , an Apache-2.0 licensed Go CLI for discovering, planning, migrating, and verifying Ubuntu and Debian servers. The rule I would not compromise on The source server must remain read-only. HostShift does not install packages, stop services, enable maintenance mode, create temporary archives, or change configuration on the source. It reads approved facts and streams data directly to the target. Any target mutation requires an explicit CLI apply command. What it currently covers Docker Compose projects and standalone containers MySQL/MariaDB, PostgreSQL, and Redis Nginx, Apache, Caddy, and systemd services SSH and firewall configuration PHP-FPM, Supervisor, Fail2ban, Certbot, and Logrotate Migration planning, audit journals, status, resume, rollback metadata, and verification checks The migration engine is deterministic Go code and does not need AI. I also added an optional Codex plugin and a deliberately non-apply MCP interface for discovery, planning, review, and dry runs. Actual changes stay in the human-operated CLI. Testing real migrations I did not want to call it tested just because a few unit tests passed. The repository includes Docker migration matrices and real Lima VM matrices covering Ubuntu 22.04, Ubuntu 24.04, Ubuntu 25.10, Debian 12, and Debian 13, including cross-distribution moves. The VM tests also reboot the target and verify persistence while comparing source snapshots before and after the migration. The project is still new, so I expect real-world edge cases. I am sharing it now because feedback from people who actually move and maintain servers will be more useful than polishing it alone forever. GitHub: https://github.com/oguzhankrcb/HostShift Documentation: https://hostshift.karacabay.com

AWS released an official Agent Toolkit that plugs Claude Code, Codex, and Cursor directly into your AWS account through a single MCP server. Instead of wiring up IAM roles and endpoints by hand, you install one plugin and the agent can search AWS docs, run sandboxed Python, and follow curated cloud skills with full CloudTrail audit logging. What the AWS Agent Toolkit Actually Is The Agent Toolkit for AWS is an open-source project (published on GitHub at aws/agent-toolkit-for-aws ) that bundles two things: The AWS MCP Server — a managed server that gives agents access to AWS through the Model Context Protocol. Agents can search AWS documentation and pull service information without authentication. To actually execute AWS API calls, run Python in a sandboxed environment, or follow curated skills, the agent authenticates through your existing IAM credentials. Agent plugins — single-install packages that bundle the MCP server configuration and a curated set of agent skills, so you don't configure endpoints and install skills one by one. The point is consolidation. One endpoint, IAM-based access controls, CloudWatch metrics, and CloudTrail logging of every API call for audit visibility. Which Agents It Supports Per AWS's own documentation, plugins ship for Claude Code, Codex, and Cursor . Kiro connects to the AWS MCP Server directly without needing a plugin, and any MCP-capable agent can point at the server manually. The install flow is refreshingly short for Claude Code: /plugin install aws-core@claude-plugins-official /reload-plugins The aws-core plugin is the recommended default — it bundles the MCP server config and skills covering service selection, infrastructure as code (CDK and CloudFormation), serverless, containers, storage, observability, billing, SDK usage, and deployment. A second plugin, aws-agents , provides additional agent-oriented capabilities. Why This Matters for Coding-Agent Users Until now, getting an agent to safely touch your cloud account meant h

This is a submission for Weekend Challenge: Passion Edition What I Built Every developer has a graveyard of side projects — started with fire, abandoned quietly on a Tuesday. They deserved better than an empty GitHub repo gathering digital dust. DevGraveyard is a gothic memorial platform where developers give their abandoned passion projects a proper burial. Connect your GitHub, pick a dead repo, carve its epitaph — and watch Gemini AI write a dramatic breakup letter from you to the project. Here's what it does: ⚰️ Bury a project — 3-step burial wizard: pick a repo → choose cause of death ( "Never Made it Past Localhost" , "Ran Out of Weekend" , "It Was Complicated" ...) → write an epitaph 🪦 Real tombstone data — pulls your actual commit history: peak obsession streak, most commits in a single day, last commit message ( your final words ) 🤖 AI Eulogy — Google Gemini writes a dramatic breakup letter from you to the project, referencing your real commit data 🕯️ Community mourning — light candles, leave RIP messages, vote to resurrect projects 🌐 3D Graveyard — a full Three.js interactive cemetery: bare trees, fireflies, flickering candles, soul wisps, resurrection pulse rings. Click any tombstone to interact My own ARweave repo had 56 commits, a 2-day peak streak, 30 commits on its best day. Cause of death: "Never Made it Past Localhost." Last words: "feat: overlay plane in 3D builder — drag/scale image on marker, position saved to DB and restored in AR viewer." It worked until it worked. Demo 🔗 Live → devgraveyard.varshithvhegde.in Code Varshithvhegde / devgraveyard Give your abandoned passion projects a proper burial. A gothic graveyard for dead side projects. ⚰️ DevGraveyard A memorial for your abandoned side projects. They deserved better than an empty GitHub repo gathering digital dust. Live → devgraveyard.varshithvhegde.in What is this? Every developer has a graveyard of passion projects — started with fire, abandoned quietly on a Tuesday. DevGraveyard gives them

Most things that ship under "browser MCP" are the same thing wearing different names: an autonomous agent with a do-anything tool, pointed at your browser, told to figure it out. The pitch is capability. The unspoken cost is that a runtime which can do anything can be steered into doing anything. We just published the opposite, and we published it in the open. github.com/e2llm/e2llm-sifr is now the canonical home for SiFR - the format spec, the taxonomy, the MCP server manifest, real page captures, per-client configs, and the model skill. MIT-licensed. The capture engine and the server stay a hosted product; the format and the interface are open. This post is about why that split is the whole point. E2LLM is not an agent This comes first because everything else follows from it. An agent decides and acts on its own. It plans, it loops, it takes steps toward a goal with you out of the path. That autonomy is the feature - and it is also the attack surface. A runtime that can do anything is a runtime that can be talked into anything. E2LLM is a perception layer, not an agent. It gives whatever model you already use senses for the browser: structured sight, and a small set of narrow, individually-gated actuators. It does not plan, does not loop, does not decide. Your model does the reasoning. E2LLM reports what a page is and carries out one explicit instruction at a time. Nothing runs while you look away. Perception substrate versus autonomous runtime. That line is the design, not a disclaimer on top of it. What SiFR is - and the three things it isn't SiFR (Salience-Indexed Flat Relations) is the capture format at the center of E2LLM. From a distance it can look like a tidy DOM dump or an accessibility tree. Mechanically it is neither, and the difference is the entire value. Not a DOM dump. A dump serializes the tree as-is: everything, in document order, noise included. SiFR selects and ranks. It scores every node by salience, drops scaffolding, and flattens the survivor

The first time a client said yes to a quote without hesitating, I felt sick. This was early on. I'd sent over a rate for a batch of articles, my palms were actually sweaty over the email, and the reply came back in under an hour. "Sounds great, when can you start?" No pushback, no negotiation, nothing. I should have been thrilled. Instead, I sat there doing the math on how much more I could have charged, and I knew, the way you just know sometimes, that I'd priced it too low. His enthusiasm was the tell. That queasy feeling taught me more than any pricing guide ever did. If a client says yes instantly and happily, you were cheap. I've been freelancing for about eight years now, all of it writing and content work, most of it solo from a spare room in my house. I've priced my work a dozen different ways over that stretch, and I've gotten most of them wrong at some point. So here's what I actually believe about pricing, after enough scars to have earned an opinion. Per-word pricing quietly punishes you for getting better I started out charging by the word, like a lot of writers do. Five cents a word, sometimes six if I was feeling brave. It felt safe because it was easy to explain and easy for a client to say yes to. A 1,500-word article costs this much. Clean. Predictable. The problem showed up slowly. The better I got, the worse that model treated me. Early on I'd pad a piece to hit a word count because more words meant more money, which is a genuinely insane thing to be incentivized toward as a writer. Then I spent years learning to cut. Learning that the sharpest version of an article is usually the shortest one that still does the job. And every ounce of that hard-won skill made me poorer, because a tight 900-word piece that took real judgment to shape paid less than a bloated 1,400-word one I could have written half-asleep. Think about how backwards that is. I was being paid the least for the writing I was proudest of. The stuff that took a decade to be able to d

Engineering prompts inside your source code is a recipe for deployment fatigue. If you've spent any time moving an AI feature from a prototype to production, you know the specific frustration of 'prompt drift.' You make a subtle tweak to a system instruction—perhaps changing how the model handles edge cases in JSON formatting—and suddenly you're forced into a full CI/CD cycle. A PR, a review, a build, and a deployment, all because of three words changed in a long string constant. In a mature engineering organization, your application logic should be decoupled from your prompt instructions. The code handles the orchestration, the plumbing, and the security; the prompts represent the dynamic configuration. This is what LLMOps aims to achieve, but until recently, there was a massive friction gap between managing these prompts in a dashboard and actually using them inside an agentic workflow. This is where the Humanloop MCP server changes the interaction model entirely. It's not just about having a central repository for strings; it's about bringing those strings into your execution context—your IDE, your Claude instance, or your Cursor agent—as actionable tools. The Architecture of Prompt-as-a-Service The core idea here is treating prompts as versioned assets rather than hardcoded constants. By using the Humanloop API via MCP, you're essentially turning prompt management into a service call. When I look at the toolset available in this server, the first thing that stands out isn't just the ability to read data—it's the ability to manipulate state. Take upsert_prompt for instance. You aren't just fetching text; you can create or update configurations directly from your agent. This transforms your development loop. Instead of context-switching between a browser tab with Humanloop and a terminal, you can instruct an agent to 'Refine the customer-support-reply prompt to be more concise and save it.' The agent performs the engineering work and updates the source of truth in

This is a submission for Weekend Challenge: Passion Edition What I Built Heirloom AI preserves family recipes & skills using Gemini multimodal AI. Upload handwritten cards, cooking photos, voice recordings, or gesture videos — it generates structured archive entries with poetic memory cards, evidence-based ingredients (with confidence levels), physical-cue step guides, and prominently flagged "speculative gaps" for family verification. Includes AI illustration generation and image enhancement. Full-stack React + Express app, persists in localStorage, exports Markdown. Demo heirloom-ai.ai.studio Code gxobst / Heirloom-AI Transform messy family recipe cards, verbal kitchen instructions, or raw cooking photos into beautifully structured, editable archive entries with Gemini. Heirloom AI 🌾 Preserve the recipes, rituals, and skills that live in family memory. Heirloom AI is a web application designed to transform messy personal materials—such as raw photographs, chaotic scribbles, handwritten notes, verbal stories, videos, and voice recordings—into beautiful, structured, editable, and shareable archival entries. The first MVP focuses heavily on family recipes , because culinary traditions are deeply emotional, practical, highly visual, and uniquely susceptible to being lost across generations. 📖 What It Does Rather than generating generic, standardized recipes off the web, Heirloom AI acts as a warm oral-history and preservation assistant. It processes your specific memory context and images to draft custom archives containing: Evocative Memory Cards : A warm, poetic summary and quote summarizing the tradition. Personal Narrative & Lore : Captures the emotional voice and regional background. Evidence-based Ingredients : Checklists tracking what ingredients are visible or described. Physical Cue Guides : Instruction steps… View on GitHub How I Built It Single Node server running Express + Vite SPA (no CORS issues). Two Gemini models: gemini-3.5-flash with schema-constrain

✓ Human-authored analysis; AI used for formatting and proofreading. Delivery teams have DORA. Four metrics — deployment frequency, lead time for changes, mean time to restore, change failure rate that predict whether a team is shipping well. Thoughtworks recently added a fifth: rework rate, measuring how much of the pipeline is consumed by fixing work previously considered complete. These metrics changed how delivery organizations operate. Because they're leading indicators. They tell you the trajectory before the outcome arrives. A team with increasing lead times is heading for trouble. A team with rising rework rate is accumulating debt. You see it in the metrics before you see it in the incidents. Security teams have no equivalent. What security teams measure today Finding counts. "We found 247 misconfigurations this quarter." More scanning produces more findings. A team that scans more frequently or adds a new tool sees the number go up which looks worse even if posture is improving. Finding counts measure scanning effort, not security health. Compliance percentages. "We're 94% compliant with CIS Benchmarks." This measures the last audit, not the current trajectory. A team at 94% today might be at 87% next week if three Terraform changes introduced misconfigurations. The percentage is a snapshot, not a trend. It rewards breadth of coverage over depth. 94% across 200 checks sounds better than 100% across 50 checks, even if the 50 are the ones that matter. Incident counts. "We had two security incidents this quarter." This is a trailing indicator. It measures failures that already happened. A team with zero incidents might have excellent posture or might have excellent luck. You can't tell. By the time the count goes up, the damage is done. None of these answer the question delivery teams answer with DORA: are we getting better, and how fast? The mapping The five DORA metrics adapt directly to security posture. The definitions are concrete and measurable from eval

For 4 years, I lived in the world of IT Operations. My days were spent handling incident response, managing data lifecycles, and making sure systems stayed online. I learned how to troubleshoot under pressure, talk to frustrated users, and keep the business running. But I had a lingering frustration: I was always fixing other people's code. I never got to build it. And more importantly, I was fixing problems manually that I knew could be automated. So, I decided to make a massive pivot. I went back to university (VILNIUS TECH) and recently started a Java Engineering internship at Coherent Solutions. My goal isn't just to become a Java developer. My goal is to bridge the gap between Development and Operations- DevOps . In my first few weeks at Coherent, we started learning about enterprise architecture. But the moment that truly clicked for me was when I built my first Docker image for our project. In my past IT life, deploying an app was a nightmare. "It works on my machine!" was a constant joke (and a constant headache for the Ops team). Setting up environments, installing the right Java version, configuring databases—it was manual, error-prone, and boring. Then I wrote a Dockerfile . I packaged our Java application and its dependencies into a single, isolated container. Suddenly, I realized: This is how you solve the "works on my machine" problem forever. As someone who used to be the guy manually fixing those environment issues, writing a few lines of code to completely automate that process felt like a superpower. I'm starting this blog to document my journey in real-time. I'm currently diving deep into: 🔹 Java 21 (the newest LTS—highly recommend checking out Virtual Threads!) 🔹 Spring Boot & enterprise backend architecture 🔹 Docker & containerization 🔹 Next up: CI/CD pipelines and Infrastructure as Code (Terraform) If you are currently stuck in IT Support or SysAdmin roles and dreaming of becoming a DevOps or Software Engineer—you aren't alone. Let's learn toge

Intro An MCP server can look like another API layer: expose existing REST endpoints as tools and call it a day. Both receive input, execute backend logic, and return a result. But they operate under different assumptions. This two-part series explains why directly wrapping REST APIs is a bad default. This first article covers the differences in their runtime environments. The second will discuss how those differences should affect MCP design (you already know how to design a good REST API ). We can see those differences more clearly by comparing the two across several dimensions. Dimensions The consumer With REST, developers encode control in application logic. The application knows when to call an endpoint, what arguments to send, and how to handle the response. Those decisions are made during development. With MCP tools, much of that control moves to the AI agent. The model interprets the request, chooses a tool, constructs its arguments, evaluates the result, and decides what to do next. The harness can restrict it, but the model is still part of the control flow. A REST client already knows why it is making a call. An agent must first decide whether a tool is relevant at all. MCP tools The context A REST application can draw from application state, cookies, memory, and user input. Code written by a developer determines which parts become request parameters. An agent can draw from the current request, conversation history, and previous tool results. The MCP server does not see this context automatically, but the model may turn parts of it into tool arguments at runtime. The difference is who selects what reaches the backend: predetermined code or a model reasoning over a changing conversation. The action model REST APIs tend to expose focused, fine-grained operations that application code can compose. Keeping endpoints simple and stable limits regressions because a developer has already written and tested the workflow that connects them. With MCP, the agent often

Playing the game of writing technical post was funny two years ago. Now, I am a mod and it is...

This is The Stepback, a weekly newsletter breaking down one essential story from the tech world. For more on the data center buildout, follow Emma Roth. The Stepback arrives in our subscribers' inboxes on Sunday at 8AM ET. Opt in for The Stepback here. How it started Years before the AI boom threatened local power […]