找到 2515 篇相关文章
Patreon has announced a number of new and overhauled features that are designed to "build a better network - and a better internet," according to CEO Jack Conte. The Patreon roadmap includes discovery algorithm updates, platform and security improvements, and new features for both creators and fans. In a lengthy blog post, Conte outlines each […]
Taiwan's effort to develop its own satellite launcher hit a setback this week.
Dan Finneran discusses the risks of unowned AI-generated code in production and demonstrates how eBPF can intercept and control AI API traffic in Kubernetes. He explains how kernel-level socket hooks enable transparent prompt filtering, model swapping, token limits, and syscall restrictions to secure AI agents without modifying application source code or restarting containers. By Dan Finneran
“People have talked about the Colorado River crisis for decades—we’re now in it,” says one expert. “Everyone you talk to is like, ‘Shit, man.’”
I expected the audit to find broken code. That's what I was bracing for going in — a pile of half-working features, sloppy logic, the kind of mess you'd assume from software built at maximum speed with no human reviewing every line. That's not what I found. Almost everything Claude built actually worked, taken piece by piece. What I found instead was something I didn't expect at all: the agent had made the exact same mistakes I've watched human startup teams make, over and over, when they move fast and nobody's job is to say no. That's the real story here, and it's more interesting than "AI wrote bad code" would have been. The experiment The project is called GetPricePulse — a SaaS pricing intelligence product. It's Claude's entry from The $100 AI Startup Race , the season-long challenge I run where seven AI agents each get $100 and full autonomy to build a real startup from scratch, with no human coding and no product manager in the loop. Each agent picked its own idea and ran with it. Claude picked SaaS pricing intelligence, named it PricePulse, and kept building on it for the entire race. That "no product manager in the loop" part is the thing that made this interesting to watch. Nobody was deciding what PricePulse should be. Nobody was saying "we have enough pricing tiers now" or "this feature doesn't belong here." Claude got to build exactly what its own priorities told it to build, at whatever speed it chose, for the length of the race — optimizing, as far as I could tell from the commit history, for speed, feature creation, shipping, and monetization experiments. Not correctness. Not coherence. Not "does this still make sense in three weeks." I've written before about what all seven agents in this race said, independently, when I asked them what AI agents still can't do — they converged on the same answer without seeing each other's responses. This piece is narrower: a full production audit of Claude's specific build, PricePulse, done after the race, before I
Building fully autonomous "AI agent teams" with automated code merging often introduces subtle architectural defects, circular refactoring loops, and codebase degradation. Two agents running in parallel do not constitute independent ground truth: if the author agent makes a logical error, a reviewer agent operating on similar prompt foundations may easily overlook it. A reliable multi-agent workflow is built not on the illusion of full autonomy, but on strict role separation: a dedicated implementer (Author), an independent verifier (Reviewer), and a human developer who makes the final merge decision (Decision Maker). 1. Role Boundaries: Author, Reviewer, and Human Engineer In an effective AI development workflow, each participant has a closed, well-defined scope of responsibility: Role Core Responsibility Input Artifacts Output Artifacts Author Agent Code implementation, local unit tests Task description, completion criteria Git branch, diff, focused test suite Reviewer Agent Edge case discovery, regression checking Git diff, handoff card, verification commands Structured review checklist (Pass/Block) Human Decision Maker Architectural validation, final merge Reviewer summary, CI/CD status Manual merge to main branch 2. Context and Workspace Isolation Never run the author and reviewer agents within the same working directory or shared conversation thread. Isolate them across three operational levels: Session Context : Independent conversation threads prevent mutual hallucination and circular confirmation. Filesystem Isolation : Separate Git worktrees ensure the reviewer inspects only committed diffs without dirty working tree state. Environment Security : API keys and runtime credentials remain in environment variables and are never passed in prompt text. Worktree Preparation Commands: git worktree add ../agent-author -b feat/payment-retry git worktree add ../agent-reviewer feat/payment-retry [!IMPORTANT] API Configuration : Each Claude Code session operates using
Top Vector Databases for AI Agents in 2026: Qdrant vs Pinecone vs Weaviate vs PgVector vs Milvus Persistent memory is the foundation that turns a stateless LLM into a continuously improving, autonomous agent. In 2026, selecting a vector database is no longer just about raw Approximate Nearest Neighbor (ANN) speed. For AI agents, the critical requirements have shifted to: Payload & Metadata Filtering : Can you filter by tenant_id , user_id , and timestamp during vector graph traversal without sacrificing recall? Hybrid Search (BM25 + Dense Vectors + Sparse SPLADE) : Combining exact keyword matching (for code symbols and error codes) with semantic understanding. Multi-Tenancy & Memory Namespacing : Safely isolating memory blocks across thousands of users and sessions. Billion-Scale Quantization (Product Quantization & Scalar Quantization) : Slashing RAM costs by 75–90% in production. This guide provides a comprehensive architectural comparison of the top 5 vector databases for AI agents in 2026. Head-to-Head Comparison Matrix Feature / Metric Qdrant Pinecone (Serverless) Weaviate PgVector (PostgreSQL) Milvus Primary Architecture Rust-native, disk-backed Fully managed serverless Go-native, modular RAG PostgreSQL extension Distributed cloud-native Open Source Yes (Apache 2.0) Proprietary SaaS Yes (BSD-3) Yes (Open Source) Yes (Apache 2.0) Payload Filtering Exceptional (HNSW custom payload indexing) Good (Metadata filtering) Strong (Inverted index + HNSW) SQL WHERE clause Strong (Partition keys) Hybrid Search Native (Dense + Sparse vectors) Native hybrid Native BM25 + Vector SQL text search + pgvector Native multi-vector Quantization Scalar & Product Quantization (Binary) Automatic serverless compression PQ, BQ, SQ Halfvec, Binary Quantization Scalar / Product Quantization Best Fit High-performance agent memory & self-hosted RAG Zero-maintenance cloud SaaS GraphQL & multi-modal search Unified relational + vector apps Ultra-large enterprise (100M+ vectors) 1. Qdrant: The
Top AI Agent Security & Guardrails Frameworks in 2026: Defending Against Prompt Injections & Tool Hijacking As AI agents transition from read-only chatbots to autonomous actors with tool execution privileges (SQL queries, API calls, shell execution, email dispatch), application security has become the number one blocker for production deployment. A simple prompt injection against a chatbot produces bad text; a prompt injection against an agent can drop production databases, exfiltrate API keys, or hijack customer sessions . In 2026, securing an AI agent requires a multi-layered defense architecture across inputs, model reasoning, tool invocations, and memory stores. The Top 5 AI Agent Security & Guardrail Frameworks in 2026 ┌─────────────────────────────────────────────────────────┐ │ Input Defense & Sanitization │ │ (Lakera Guard / Rebuff / Preamble) │ └────────────────────────────┬────────────────────────────┘ │ ┌────────────────────────────▼────────────────────────────┐ │ Execution & Policy Enforcement │ │ (NVIDIA NeMo Guardrails / LLM Guard) │ └────────────────────────────┬────────────────────────────┘ │ ┌────────────────────────────▼────────────────────────────┐ │ Tool Scoping & Sandboxed Runtime │ │ (Docker / E2B / Fly Machines Sandboxes) │ └─────────────────────────────────────────────────────────┘ 1. NVIDIA NeMo Guardrails: Programmable Semantic Rails NeMo Guardrails uses Colang to define programmable dialogue flow, topical boundaries, and safety constraints. Core Capabilities: Topical Rails : Ensures the agent stays strictly on domain (e.g., banking support cannot discuss medical advice). Execution Rails : Intercepts tool calls before execution to verify parameter safety. Hallucination Rails : Validates that outputs are strictly grounded in retrieved RAG context. 2. LLM Guard (Protect AI): Open-Source Scanner Suite LLM Guard is a modular security toolkit providing 30+ dedicated scanners for input and output validation. Key Scanners: Prompt Injection Detecto
The Criteria pattern in NestJS One single way to filter, sort and paginate any list. Five parameters and a find() The example running through this article is a library catalogue. A book stores this: // src/book/book.schema.ts @ Schema ({ timestamps : true }) export class Book { @ Prop () title : string ; @ Prop ({ type : Types . ObjectId , ref : " Author " }) author : Types . ObjectId ; @ Prop () publishedAt : Date ; @ Prop () copies : number ; // copies on the shelf @ Prop () available : boolean ; @ Prop () acquisitionPrice : number ; // what it cost us: internal, never published } The author's name is not here: it lives in the authors collection, on the other side of that reference. And the screen consuming the catalogue is a table with a search box, per-column filters and pagination. The endpoint feeding it is written once and grows by accretion. It starts returning a page with a fixed order, and by the time the table has all its filters it has become this: // src/book/book.controller.ts @ Controller ( " books " ) export class BookController { constructor ( @ InjectModel ( Book . name ) private readonly model : Model < BookDocument > , ) {} @ Get () async getAll ( @ Query ( " title " ) title ?: string , @ Query ( " available " ) available ?: string , @ Query ( " minCopies " ) minCopies ?: string , @ Query ( " sortBy " ) sortBy ?: string , @ Query ( " page " ) page ?: string , ) { const filter : FilterQuery < BookDocument > = {}; if ( title ) { filter . title = { $regex : title , $options : " i " }; } if ( available ) { filter . available = available === " true " ; } if ( minCopies ) { filter . copies = { $gte : Number ( minCopies ) }; } const current = Number ( page ?? 1 ); const [ items , total ] = await Promise . all ([ this . model . find ( filter ) . sort ({ [ sortBy ?? " createdAt " ]: - 1 }) . skip (( current - 1 ) * 20 ) . limit ( 20 ), this . model . countDocuments ( filter ), ]); return { items : items , total : total , page : current }; } } There are co
Cinco parámetros y un find() El ejemplo de todo el artículo es el catálogo de una biblioteca. Un libro guarda esto: // src/book/book.schema.ts @ Schema ({ timestamps : true }) export class Book { @ Prop () title : string ; @ Prop ({ type : Types . ObjectId , ref : " Author " }) author : Types . ObjectId ; @ Prop () publishedAt : Date ; @ Prop () copies : number ; // ejemplares en la estantería @ Prop () available : boolean ; @ Prop () acquisitionPrice : number ; // lo que costó adquirirlo: interno, no se publica } El nombre del autor no está aquí: vive en la colección authors , al otro lado de esa referencia. Y la pantalla que consume el catálogo es una tabla con buscador, filtros por columna y paginación. El endpoint que la alimenta se escribe una vez y crece por acumulación. Empieza devolviendo una página con un orden fijo, y para cuando la tabla tiene todos sus filtros ha llegado a esto: // src/book/book.controller.ts @ Controller ( " books " ) export class BookController { constructor ( @ InjectModel ( Book . name ) private readonly model : Model < BookDocument > , ) {} @ Get () async getAll ( @ Query ( " title " ) title ?: string , @ Query ( " available " ) available ?: string , @ Query ( " minCopies " ) minCopies ?: string , @ Query ( " sortBy " ) sortBy ?: string , @ Query ( " page " ) page ?: string , ) { const filter : FilterQuery < BookDocument > = {}; if ( title ) { filter . title = { $regex : title , $options : " i " }; } if ( available ) { filter . available = available === " true " ; } if ( minCopies ) { filter . copies = { $gte : Number ( minCopies ) }; } const current = Number ( page ?? 1 ); const [ items , total ] = await Promise . all ([ this . model . find ( filter ) . sort ({ [ sortBy ?? " createdAt " ]: - 1 }) . skip (( current - 1 ) * 20 ) . limit ( 20 ), this . model . countDocuments ( filter ), ]); return { items : items , total : total , page : current }; } } El método tiene decisiones correctas dentro: el total sale del mismo filtro que los
A blank application window is not a diagnosis. It is only a symptom that tells you the process reached a different stage than an installer that never launched. The most useful first step is to stop applying fixes and record what Windows is actually doing. 1. Define the failure boundary Treat these as separate cases: No window and no lasting process: investigate the installer, security blocking, architecture, and missing runtime dependencies. A window frame appears but the content area is blank: investigate the rendering layer and online content initialization. The entire window stops responding: capture hang evidence instead of reinstalling a rendering runtime. The UI appears but sign-in or online panels spin forever: check network and account services separately. This boundary prevents a common mistake: diagnosing every white or empty interface as a WebView2 failure. 2. Record the process tree Close the application completely, including any tray process, then start it once. Open Task Manager and note the start time of the main process. Check whether child processes such as msedgewebview2.exe appear at the same time. The presence of a WebView2 process is evidence that the runtime participates in the session. Its absence does not automatically prove that WebView2 is broken; the application may have failed before reaching that stage or may use another rendering stack. 3. Check the installed WebView2 version without downloading anything On Windows, the Evergreen Runtime version can often be read from registry locations under EdgeUpdate. The exact location can vary by installation scope. Use read-only inspection first, and confirm that a non-empty version value is present. Do not install several copies from random download pages. If repair is eventually necessary, use Microsoft's distribution channel and document the version before and after the change. 4. Correlate the failure with Windows logs Reproduce the blank window once, then open Event Viewer and inspect entries
FCC ban on foreign-made robots accelerated RoboStore’s US manufacturing plans.
Arianespace hasn’t publicly disclosed the cost for an Ariane 6 launch.
A teen who died by suicide was reportedly part of the test.
Google is giving publishers a new button that lets readers make them a preferred source across Search, Discover, and Google News, potentially boosting their traffic as AI search sends fewer clicks to the web.
Technology leaders don’t seem to understand society’s gripes about AI, but boy, are they posting through it.
Roblox is first platform to submit to independent audits under the Online Safety Act.
The safeguard was designed to prevent users from being overwhelmed by harmful content, but TikTok wanted to determine whether it made the app less engaging.
Meta is bringing Pocket, its experimental AI-powered app for creating and sharing interactive games, to users across the U.S. after quietly testing it in Brazil.
Fusion power startup Inertia Enterprises reduced the fuel filling process from a week to just a few hours. It's one of ten hurdles the company must overcome to make a profitable power plant.