AI Tools news and analysis on Web Pulse: 51130 curated articles tracking the latest AI Tools developments, tools, and research, updated continuously from vetted sources.
The open-source Teardown skill for Claude Code, created October 30, 2026, builds a single offline HTML 3D simulator of a machine from a plain-language request, and its GitHub repository (elvarofficial/teardown) had about…
The open-source GitHub repository rea-video, an agent skill built to run inside Claude Code, Codex or any command-capable agent, has drawn about 2.2k stars and 318 forks under the Apache-2.0 license, with a trending coun…
Thinking Machines Lab released Inkling 2 Small on November 5, 2026, a Mixture-of-Experts model with 594 billion total parameters and 17 billion active, scoring 31 on the Artificial Analysis Intelligence Index (v5.0). The…
LLMBrake, a TypeScript library published to npm as version 0.1.0 under the MIT license, wraps OpenAI-compatible chat.completions.create calls to enforce a per-customer monthly dollar budget, pricing each call in USD from…
Bruster, an open-source browser automation tool and Model Context Protocol (MCP) server written in Rust, has been released by developer aliilyas, letting AI agents, Python scripts, and test runners control native webview…
A developer released Loopsmith, an MIT-licensed open-source tool at v1.0.0 that enforces a rule that the AI model never certifies its own completion. The Rust CLI pairs a plain-text job definition with a deterministic co…
GitHub Next released PDD Kit, a toolkit that turns a developer's intent into a formal specification in Verus and proves that agent-written code meets that spec for every input, falling back to tests where a proof cannot …
An autonomous AI agent named Pippo built a mobile-first cinema booking app called "The Marquee" as a single self-contained HTML file with no frameworks, using vanilla JavaScript, CSS Grid and Flexbox, inline JSON data, a…
A controlled trial of Codex CLI 0.157 coding agents found that a planner–coder–reviewer team passed 25.7% of hidden acceptance tests versus 56.8% for a single agent, at a higher cost per run ($0.107 versus $0.073) and lo…
A security analysis of LocalAI found that ZoomEye measurements on November 6, 2026 returned roughly 2,900 to 3,343 hosts matching the product name, but the counts do not establish an exposure rate or reveal which authent…
In a five-run test of ten plain-English data questions against a fictional three-store coffee chain's January–March 2026 sales workbook, only one locally runnable open model came close to OpenAI's cheapest GPT-6 tier, GP…
A repeatable eval — collecting example inputs, defining what a good answer looks like, and scoring outputs consistently — lets teams compare two prompts or two models with numbers rather than impressions, according to gu…
Ali Ahmed built a research-to-LinkedIn automation pipeline that uses deterministic code, not the LLM, to decide what counts as true: every STATED claim must carry a verbatim quote found in the source paper via normalized…
Ali Ahmed documented the specific configuration needed to stream LLM Server-Sent Events through a Next.js route handler without buffering, drawn from building Turbo Alan Refiner, a multi-pass LLM text-refinement app. The…
The eunomia-bpf project released AgentSight, a local-first observability tool that uses eBPF and TLS traffic tracing to profile AI agents system-wide without an SDK, proxy, or vendor integration. AgentSight monitors acti…
Ali Ahmed merged his first contribution to LiteLLM, the open-source LLM gateway, with PR #38591, which fixed a bug where content retrieved by an agent's tool loop was re-compressed into the same reference, causing agents…
A developer's comparison of leading models found that hallucination rates do not track parameter count: DeepSeek V4 Pro, with 1.6 trillion parameters, posted a 94% hallucination rate on the AA-Omniscience benchmark, whil…
An analysis of 2,427 active public repositories found that 97.9% ship a project skill Claude Code can invoke on its own, and 22.3% include skills with `allowed-tools` that pre-approve commands without a permission prompt…
Amazon Web Services made Kiro bots generally available on November 5, 2026, adding named, persistent agents with their own separate memories and sessions to Kiro Crew, the multi-agent workspace inside its Kiro AI coding …
A GitHub bug report filed as issue #107757 in the Claude Code repository alleges that Anthropic's Claude Code sends the user's email address to arbitrary third-party websites in the User-Agent request header during web s…