AI 资讯

CLI that turns plain-English into real browser tests Discussion | Link

Spent the week breathing new life into DevNotion—59 commits and over 10,000 lines of code later,...

Amflow, the e-bike brand spun out of DJI, just announced its TL series, a do-it-all "eSUV" suitable for both bikepacking adventures and dropping the kid at daycare on your cycle to work. The all-terrain TL series is built around Amflow's incredibly compact yet powerful Avinox M2 mid-drive motor. The Amflow TL Carbon offers 125Nm of […]

The easiest way to misunderstand LangGraph is to see it as “LangChain, but with more steps.” That misses the point. LangGraph becomes useful when an agent is no longer a single prompt or a simple chain. It becomes useful when the workflow has state, branches, tool calls, human approval, checkpointing, and recovery behavior that must be inspected before the agent is trusted inside a real AI host. I used the Doramagic LangGraph manual as the source-backed reading layer for this note: https://doramagic.ai/en/projects/langgraph/manual/ This is an independent project guide, not an official LangGraph document. I use it as a pre-adoption checklist: what should be understood before wiring a project into Claude, ChatGPT, Cursor, Codex, or another AI host. The point is not to create another prompt library. The useful artifact is a capability resource pack: a manual, source map, boundary notes, pitfall log, smoke check, lightweight eval criteria, feedback notes, and host-ready context that help a developer decide what to verify before adoption. 1. The real boundary is State, not the prompt For a one-shot model call, the prompt is often the main boundary. For LangGraph, the first boundary is the State schema: which fields move between nodes; which fields a node may update; how concurrent branches merge values; which values enter a checkpoint; which values should never be persisted. This is why reducers matter. A message list is usually not just overwritten. It needs an append or merge rule such as add_messages or the TypeScript equivalent. That small implementation detail decides whether parallel work preserves context or silently drops it. My preferred first run is not a “universal agent.” It is a tiny graph with one State schema, one node, one partial update, and one explicit reducer. If that is not clear, adding tools will only hide the problem. 2. compile() is the boundary between description and runtime Before compile() , a LangGraph graph is a description: nodes, edges, c

Payment data pipelines fail in ways that ruin a payments engineer’s week, and the failures rhyme. The dashboards froze. Fraud scores arrived after the transaction had already cleared. Settlement reports came in stale. Nobody slept. The frustrating part is that the same data architecture had run fine for years. So, what changed? The honest answer is that batch thinking does not survive contact with real-time payments. A lot of banks built their data foundations in an era when nightly jobs were good enough. Load the warehouse overnight, run the reports in the morning, move on. That rhythm worked when money moved slowly. It does not work when a customer expects an instant confirmation and a fraud engine has milliseconds to make a call. Here is where things crack. Real-time payment rails push a constant stream of events instead of a tidy nightly dump. Your pipeline now has to ingest, transform, and serve data while transactions are still happening. Add ISO 20022 into the mix and the pressure climbs. ISO 20022 messages are rich. They carry far more structured detail than the old formats, which is wonderful for analytics and miserable for a pipeline that was never designed to parse that much context at speed. This is not a fringe concern either. Swift reported that by the time its MT/ISO 20022 coexistence period closed in November 2025, around 80% of daily traffic was already running on the ISO 20022 format, with more than 3.1 million of these messages exchanged every day. The rich-data era is the default now, not the roadmap. Then there is the fraud-scoring window. Fraud models need fresh features. Account behaviour over the last few minutes, velocity checks, device signals. If your pipeline takes thirty seconds to surface that data, the fraud decision is already too late. You are essentially detecting fraud after the loss. That gap between when data is created and when it becomes usable is the silent killer in most payment systems. And the cost of getting it wrong runs

If your LLM costs are climbing, the instinct is almost always the same: swap to a cheaper model. GPT-4 to GPT-4-mini. Claude Opus to Claude Haiku. Sometimes that helps a little. It rarely fixes the actual problem. The actual problem, in most workflows I've looked at, is that every step gets routed through the LLM, even the steps that don't need language reasoning at all. This post breaks down a simple mental model for deciding what should and shouldn't touch an LLM, with a working example you can adapt. The four components of any AI workflow Every automated workflow — whether it's a support ticket router, a fraud check, or a content pipeline — is built from some combination of four building blocks. They get treated the same once a workflow diagram is drawn flat, but they have wildly different cost and latency profiles. Component What it does Think of it as Typical cost Trigger Starts the workflow The doorbell ~$0 Deterministic ML Structured predictions — classify, score, rank The calculator Cents per 1,000 calls LLM / Generative Reads, writes, reasons in language The writer Dollars per 1,000 calls Tool / API Fetches or writes real data The hands Cents per 1,000 calls The gap between row 2 and row 3 is the whole article. A classifier and an LLM call can solve the exact same problem, but one costs roughly 100-1000x more than the other, depending on model and provider. If you're not deliberately deciding which one handles which step, you're probably defaulting to the expensive one — because in frameworks like LangChain or a quick custom agent loop, it's just easier to shove everything into a prompt. Where this actually shows up Here's a workflow I see constantly: an automated support ticket triage system. flowchart LR A[New support ticket] --> B{Classify intent} B --> C[Route to team] B --> D[Auto-draft response] D --> E[Update CRM] A naive build sends the entire ticket text to an LLM and asks it to do everything at once: classify the intent, decide routing, draft a re

A look at what's actually changing in Nepal's job market, what it means for students and working professionals, and what separates training that gets you hired from training that just gives you something to print on a resume. Nepal is at an interesting crossroads right now. On one side, the country still carries the weight of a job market that hasn't kept up with its graduates. Every year, more than 500,000 young people enter the workforce. The economy, for all its resilience, simply does not generate enough traditional jobs to absorb that number. The result is familiar to most Nepali families: children who studied hard, passed their exams, collected their certificates, and then spent months, sometimes years, waiting for something to happen. On the other side, something genuinely different is building. Nepal's IT exports crossed $1 billion in 2025, according to NASIT's estimates. Software and BPO exports grew over 20% in the first seven months of fiscal year 2024/25 alone. The government's 16th development plan has set a target of 250,000 new IT jobs and a 5% GDP contribution from the sector by 2029. International companies, from Indian IT majors to US-based outsourcing firms, are paying attention to Nepal in ways they weren't a decade ago. These two realities exist at the same time, in the same country, often in the same family. A brother driving a taxi while his younger sister lands a remote software development contract earning more than their father ever did in a government job. The difference between those two outcomes, more often than not, comes down to whether someone made the decision to learn something the market actually needs, and found a way to learn it properly. That's what this piece is about. The Skills Gap Problem Nobody Talks About Enough Nepal's IT sector is growing, but that growth comes with a problem attached: a persistent, widening mismatch between what employers need and what most fresh graduates can actually do on day one. Companies like Deer

I’ve been shipping code since before most people even knew what Git was. I've seen entire architectures built around point-to-point API integrations that were beautiful for a quarter, and then became unmaintainable monoliths by the second year. If you spend any time in enterprise software development—especially anything touching customer data or HR pipelines—you run into integration hell. The modern AI agent promises to be this universal connective tissue, right? It sounds simple enough: give it access, and boom, productivity magic. But let’s be real about what that means under the hood. When an LLM is given a tool schema, how does it get data from five wildly different systems—Salesforce for contacts, Workday for employees, Zendesk for tickets, Greenhouse for candidates? The naive approach, and frankly, most teams still take it this way, is to build bespoke orchestration services. You create a microservice that accepts an input query (e.g., 'What did Jane do last month?') and then contains specialized logic: if the name format looks like a CRM record, call salesforce_api ; if it sounds HR-related, hit workday_endpoint , etc. This is debt acceleration disguised as architecture. You are not building an integration layer; you are building a brittle routing table that requires human intervention every time one of the underlying APIs changes its schema or rate limit structure. It’s glue code for glue code's sake, and it has a massive maintenance overhead. The core problem is that most agents see data sources as functional silos , not integrated components of a single operational truth. Your CRM thinks about accounts; your HRIS thinks about job codes; your ATS tracks keywords. They all speak different dialects of 'person' or 'business unit.' When an agent needs to know, say, which employees (HRIS) are currently candidates in the pipeline (ATS) who also have a linked account record (CRM), you hit a wall. The solution isn't more specialized microservices. The solution is s

After using Claude Code at work for months I wanted to use it on my own projects too. Most probably because my vpn was on I got banned after 1 hour of usage. I got my 120$ back. 1 month later I signed up with vpn off. But this time probably because I used the same credit cart (and that's the only card I can use) they banned me again. Even after I contacted support I got a generic "we have determined that we cannot reinstate your account at this time due to a violation of our Usage Policy" answer

I recently gave a version of this talk at AI Engineer Europe in London. What follows is the fuller story — what we found when we looked at thousands of skills, what goes wrong, and how to fix it. You know that scene in The Matrix? Neo gets a spike in the back of his head, they upload kung fu directly into his brain, and he just... knows it. That's what a skill is for an AI coding agent. You write a markdown file — a SKILL.md — and the agent loads it when the task matches. Suddenly it knows your team's deployment process, or how your API handles pagination, or that you never use semicolons. It's not code. It's context. Procedural knowledge, injected at the right moment. The thing is — Neo's upload worked perfectly. Ours? Not always. Skills are everywhere now We spent some time analysing essentially all of public GitHub. In November last year, 12 repos had SKILL.md files. By March — five thousand four hundred and sixty. That's 450x growth in fourteen weeks. Skills went from zero to 27% of all agent config activity in three months. Faster adoption than CLAUDE.md , AGENTS.md , or any of the dotfile formats before them. And 1 in 12 merged PRs on GitHub now touches an agent config file — 8.4%, up from basically zero eighteen months ago. This is not a niche thing anymore. This is how people are working. Watch on YouTube But are they electrifying? Ninety percent of agent config files are never updated after creation. Write once, forget forever. Your codebase evolves every day. Your dependencies change. Your API contracts shift. But the instructions you gave your agent? Frozen in time. For Gemini files it's even worse — 97% are write-once. And the purpose-built "skill-as-product" repos? Over half are under 50 kilobytes. Wrapper repos. Many are AI-generated. High churn, low staying power. We have this explosion of skills, and most of them are going stale the moment they're committed. What we did about it The DevRel team at Tessl spent a couple of months doing something pretty

Author: Alex Isa (Webappski). This is the dev-tutorial cut of a longer piece on the Webappski blog — terminal-first, fewer words on the why. If a buyer asks ChatGPT "best CDN providers 2026" and your product is not in the answer, you lose the sale before you ever see the lead. The only honest way to know whether that is happening is to ask the engines the questions your buyers ask and read the raw answers — not trust a single dashboard score. Here is the loop we at Webappski run for a client, with the open-source tool aeo-platform (MIT, zero runtime deps). 1. Install and point it at the client's domain npm install -g aeo-platform cd client-audit && aeo-tracker init init writes a .aeo-tracker.json . The three things that matter: { "brand" : "Northwind CDN" , // illustrative, fictional brand "domain" : "northwind.example" , // registrable domain — subdomains count, spoof hosts don't "engines" : [ "openai" , "gemini" , "anthropic" ], // ChatGPT, Gemini, Claude "queries" : [ "best CDN providers 2026" , "best low-latency video streaming CDN 2026" , "alternatives to the market-leading CDN 2026" ] } The questions ARE the audit. A basket of vanity phrases produces a flattering, useless number; a basket of the buyer's real decision questions produces a number that predicts revenue. Freeze it, so next month's run is comparable. 2. Run it — sampled, not one noisy shot AI answers are non-deterministic: ask the same question twice and you can get a different list. A single pass turns that noise into a fake-precise number. So run each cell several times and let the score carry a confidence interval instead of pretending one shot is the truth: # plain single-shot run aeo-tracker run # sample each cell N times — the score comes back with a Wilson confidence interval aeo-tracker run --samples = 5 With --samples=5 , every (query × engine) cell is queried five times; the headline presence rate is then reported as a Wilson interval, and small samples are flagged as small rather than so

Hello Dev Community! 👋 It is officially Day 71 of my unbroken 100-day full-stack engineering run! After mastering polymorphic multi-part storage configurations yesterday, today I successfully crossed into core transactional operations: Engineering a High-Fidelity "Confirm and Pay" Checkout View and Wiring Database Inbound Array Modifications! In real-world booking platforms, processing a successful transaction requires more than updating an absolute view; you have to link documents relationally across collections. Today, I wired that entire execution pipeline together! 🧠 What I Handled on Day 71 (Checkout Engineering & Target Mutations) As displayed across my latest system files in "Screenshot (164).png" and "Screenshot (165).jpg" , handling payments runs through structured backend steps: 1. High-Fidelity Checkout Component ( /reserve ) I built out the detailed split-pane verification interface visible in "Screenshot (164).png" . The layout captures target trip date selections, total guests parameter caps, card input structures, and computes subtotal ledgers dynamically: Base Compute: $9000 x 5 nights = $45000 . Transactional Upgrades: Appending structured service charges ( $85 ) and local tax calculations ( $42 ) to update the final sum directly to $45127 . 2. Live Document Array Mutators (MongoDB User List Insertion) The most crucial logic happens when the user clicks the primary validation trigger labeled Confirm and pay : The inbound route controller extracts the targeted property identity token ( home._id ) via an embedded hidden input container. Instead of running isolation updates, it issues an atomized update operation straight into our MongoDB user records array (e.g., using Mongoose operators like $push or tracking active profiles inside our custom data state loops). This appends the exact property listing target ID directly into the user's booking history array database matrix! 🛠️ View Markup Code Integration View As showcased in my VS Code script structu

AI video & audio effects SDK for real-time apps Discussion | Link

There's a version of dogfooding that's a slogan, and a version where your own employees can't ship code until the bug is fixed. We run the second kind. The Authagonal staff console, the one we use to manage every tenant, authenticates through Authagonal itself: SAML single sign-on from our Entra directory, with SCIM deciding who gets in and what they can do. There is no separate admin password table. We deleted it. If our own SAML is broken, we are locked out of our own product. That's uncomfortable in exactly the useful way. It turns "SSO is an enterprise feature we support" into "SSO is the only way the people who built this get to work today." Here's what being our own customer caught. A trimmed build that broke signature verification We publish the auth server trimmed, to keep the image small. Trimming aggressively deletes code it can't prove is used, and reflection hides usage from it. .NET resolves its XML-signing crypto algorithms by name, reflectively, through CryptoConfig . The trimmer couldn't see those types were needed, removed them, and SignedXml quietly came back unable to build the algorithm. SAML signature verification, the step that proves the login is real, threw a null reference at runtime. The unit tests passed, because they ran against the untrimmed build where the types still existed. Only the trimmed production artifact failed, and it failed at the exact moment a human tried to sign in. We ship the auth server untrimmed now, with a healthy distrust of trimming anything near reflection-based crypto. If we only supported SAML rather than living on it, this is a customer's incident report instead of ours. Provisioning is the actual login Authenticating a user is the easy half of SSO. The hard half is deciding what they're allowed to do and keeping it in sync as people join and leave. We drive that with SCIM: Entra group membership maps to roles, resolved at the moment a token is issued, not copied once at account creation. Add someone to the righ

OrderHub Day 4: never trust the client. Today the backend gets proper Bean Validation — bad requests are rejected at the edge with a clear 400, long before they reach the business logic. And it's all declarative. ✅ Try the validating form (see the 400 body): https://dev48v.infy.uk/orderhub/day4-validation.html Three DTOs, three jobs A common beginner mistake is using one class everywhere. OrderHub keeps them separate: Request DTO ( CreateOrderRequest ) — what the client sends, and where validation lives. Domain/Entity ( Order ) — the internal model. Response DTO ( OrderResponse ) — what the API returns, so internal fields never leak. Validation is just annotations public record CreateOrderRequest ( @NotBlank @Size ( max = 120 ) @CleanText String customer , @NotBlank @Size ( max = 200 ) @CleanText String item , @Min ( 1 ) @Max ( 1000 ) int quantity ) {} Add @Valid @RequestBody on the controller and Spring checks every rule before your method runs. Break one and it throws MethodArgumentNotValidException . A custom constraint + clean errors @CleanText is a custom ConstraintValidator (rejects blank-after-trim + a small blocklist) — you can write your own rules, not just the built-ins. A @RestControllerAdvice turns validation failures into a tidy 400 with a field→message map. (Day 5 upgrades this to full RFC-7807 ProblemDetail.) 🔨 Full walkthrough (constraints → @valid → custom validator → 400 handler) on the page: https://dev48v.infy.uk/orderhub/day4-validation.html OrderHub — a production-grade Spring Boot backend, one feature a day. 🌐 https://dev48v.infy.uk · Code: https://github.com/dev48v/order-hub-from-zero

We warm your prospects before reaching out Discussion | Link

This is ymawky, a now-dynamic web server written entirely in ARM64 Assembly. I've previously posted about ymawky here: https://news.ycombinator.com/item?id=48080587 In the past month and a half, I've made some pretty major improvements: I've added CGI scripting support, so the server now supports query strings and dynamic content; and I've fully ported ymawky to run on Linux, rather than macOS-only. In addition to GET/PUT/HEAD/DELETE/OPTIONS requests, because of CGI support ymawky also accepts P