找到 1691 篇相关文章
GitHub热门项目 | Context7 Platform -- Up-to-date code documentation for LLMs and AI code editors | Stars: 59,323 | 71 stars today | 语言: TypeScript
GitHub热门项目 | Build infinite canvas apps in React with the tldraw SDK. World's best, top-most agent recommended #1 five star SDK. | Stars: 48,932 | 70 stars today | 语言: TypeScript
GitHub热门项目 | genshijin 原始人 🗿| Claude Code / Codex等AIエージェント 向け超圧縮コミュニケーションスキル。caveman の日本語版をベースに、日本語特有の冗長表現に最適化。 | Stars: 257 | 6 stars today | 语言: JavaScript
GitHub热门项目 | FastAPI framework, high performance, easy to learn, fast to code, ready for production | Stars: 100,634 | 41 stars today | 语言: Python
GitHub热门项目 | Kimi Code CLI is your next CLI agent. | Stars: 9,290 | 48 stars today | 语言: Python
GitHub热门项目 | The go-to web for your AI coding agent — local-first search, fetch, crawl & research over MCP. No API keys, no cloud, $0/query. Public beta. | Stars: 964 | 192 stars today | 语言: TypeScript
GitHub热门项目 | AirLLM 70B inference with single 4GB GPU | Stars: 23,152 | 242 stars today | 语言: Jupyter Notebook
GitHub热门项目 | LIBERATED AI CHAT | Stars: 9,367 | 63 stars today | 语言: TypeScript
As a developer, I got a bit scared and fed-up of blindly merging massive pull requests and losing context on my own codebase, so I built FluencyLoop . https://github.com/baokhang83/fluencyloop It is an open-source, local-first workflow plugin for Claude Code and Codex that ensures your code and your codebase fluency are produced together. Instead of letting an agent dump raw code, it tracks your technical familiarity locally and pauses to explain complex architectural choices or rejected alternatives only on topics you do not know yet. It also produces documentation and tracks the decisions it makes along the way. Try it out ! It feels really good to get the AI caring for my understanding 🤣
A follow-up to Part 1: the self-recall thesis — the series runs through Part 6 . Code: RE-call — everything below is measured and reproducible ( make eval ), full study in docs/ENTAILMENT_SUPERSESSION_STUDY.md . I published a thesis post about agent memory and got five comments that were better than the post. Two of them didn't just critique the design — they described, precisely, why it would fail and what would fix it. So I did the only reasonable thing: I turned both into experiments, ran them on the same eval harness the series is built on, and shipped what survived. That's RE-call v0.3 , and this post is the receipt. I want to be explicit about why I'm writing it this way. The point of publishing this series was never broadcast — it was error-correction . A design you keep in a drawer accumulates conviction; a design you publish accumulates objections , and objections are the cheapest high-quality signal you will ever get. The comment section of Part 1 did more for this codebase than any week of solo iteration. This post exists to pay that back with the thing commenters almost never receive: evidence that someone listened, measured, and changed the code. Comment 1: "A similarity score is not a confidence score" Vinicius Pereira put it in one line I've been quoting since: Proximity is a candidate; entailment is the evidence. His argument: the near-misses that hurt most are high-similarity and wrong — memos semantically adjacent to the query that don't answer it. A threshold-based gap_warning (Part 3, Part 5) waves them straight through by construction , because their similarity clears any threshold you could calibrate. The abstention signal cannot be the retriever's own score. You need a separate check that the retrieved memo actually entails an answer. He was right, and measurably so. I built a held-out challenge set of 10 near-miss queries — each names a strongly on-topic memo that does not contain the asked-for fact ("how much did the cache reduce memory usag
Over the last few months I’ve noticed a pattern emerging across AI tools. Whether it’s Claude Skills, Cursor, Codex, or custom agent frameworks, we’re increasingly giving AI agents “skills”—packages containing instructions, documentation, and sometimes scripts. The problem is… A skill is usually just a Markdown file (plus some assets). Nothing tells you: Who created it. Whether it has been modified. Whether the version your AI is executing is the same one you reviewed yesterday. Whether someone quietly injected new instructions into it. As AI agents become capable of executing increasingly powerful workflows, that becomes a real supply-chain problem. So I built Skillerr. ⸻ What is Skillerr? Skillerr is an open-source protocol and CLI that adds trust and verification to AI skills before they’re executed. Instead of treating a skill as “just another folder,” Skillerr treats it as a verifiable package. It focuses on three things. Package Integrity Every packaged skill receives a unique content-derived identifier along with cryptographic SHA-256 hashes. If any file changes after packaging—even a single character—Skillerr detects it immediately. No silent modifications. ⸻ Structured Contracts Instead of relying on long paragraphs that an AI has to interpret, a Skill contains a structured contract describing: required inputs permissions forbidden actions expected outputs whether a human has actually reviewed it This makes skills easier for both humans and AI agents to reason about. ⸻ Optional Public Provenance Authors can cryptographically sign their skills. Optionally, the package digest can also be anchored into Sigstore’s transparency log, making it independently verifiable without trusting Skillerr itself. Importantly: Only cryptographic identifiers are published. No prompts. No documentation. No knowledge base. No proprietary content. ⸻ I tried to break my own tool Before releasing it, I intentionally attacked it. First I packaged and signed a simple CSV processing s
A CSV file can preserve every byte you exported and still look different after somebody opens and saves it in a spreadsheet. The problem is not always a broken parser. Spreadsheet applications deliberately interpret text during import. Depending on the application, version, locale, and import path, that can remove leading zeros, round long integer-like values, treat E notation as scientific notation, parse compact letter-and-digit tokens as dates, or reinterpret decimal and thousands separators. I released PlainCell to make those possible interpretations reviewable before the file is opened and saved. PlainCell is a zero-dependency, local, read-only Node.js CLI and browser workspace. It reports the exact source cell, physical line, column, original text, reason, and an explicit import recipe. It does not rewrite the CSV or claim to know what the value was intended to mean. A small example Consider this export: record_id,account_code,sample,reference,amount 1,00123,JAN1,1234567890123456,"1.234,56" 2,00007,12E5,9876543210987654,"19,95" Install PlainCell from the public Codeberg npm registry: npm install --global plaincell@0.1.0 \ --registry = https://codeberg.org/api/packages/automa-tan/npm/ Then run the preflight: plaincell risky.csv The bundled fixture reports eight possible interpretations across four columns. Those findings cover several separate risks: 00123 and 00007 may lose leading zeros if imported as numbers; 16-digit integer-like references may exceed the precision preserved by ordinary spreadsheet numeric cells; 12E5 may be interpreted as scientific notation; JAN1 may be interpreted as a compact date-like token; decimal-comma and mixed grouping/decimal text depend on the chosen import locale and separators. A finding means “review this cell and the import settings,” not “data loss definitely occurred.” PlainCell cannot infer whether a value is an identifier, a measurement, a date, or a typo. Provenance instead of a generic warning “Be careful opening CSV i
variant-confidence v0.1.0: a calibrated confidence layer for variant-effect pathogenicity scores State-of-the-art variant-effect models are accurate in cross-validation but their scores are poorly calibrated on temporal data. variant-confidence adds an auditable calibration layer on top of existing predictors — it does not train a new model. The problem: accuracy is not trust Protein variant-effect predictors (AlphaMissense, ESM-1v, EVE) report pathogenicity scores, but a clinician or researcher needs to know how much to trust the number , not just its rank. The gap is calibration, not accuracy: AnnotateMissense (2026) reports MCC 0.94 in cross-validation, dropping to 0.76 on temporal ClinVar, accuracy 0.8798. A raw score near 0.9 may not mean 90% probability. Acting on an uncalibrated score is a risk. What it does variant-confidence wraps an existing predictor's score and produces a calibrated, uncertainty-aware output: Probability calibration (AC1): Platt scaling or isotonic regression over a separate holdout. Selectable, not hardcoded. Conformal prediction (AC1b): coverage 1−α intervals, split or Mondrian by gene. ECE (AC2, AC9): Expected Calibration Error reported before/after calibration, with bootstrap CI and per-bin counts. Bins with too few samples are flagged as low-reliability. Leakage-free split (AC3): temporal split by ClinVar release date with gene isolation — the same gene never appears in both train and test. This is unit-tested. Missing-score handling (AC4): works with AlphaMissense or ESM-1v alone; emits an explicit warning instead of failing silently. Non-deceptive reporting (AC7): every result includes interval/ECE + method + threshold, never a bare calibrated score. Verification (clean clone, no network) Built under a three-party governance loop: implement → independent audit in a clean clone → merge approval. ruff check . → All checks passed. pytest tests/ → 28 passed in 8.90s (offline fixture). An honest bug we caught in audit The first ECE tes
周六慢读：FROST家族的周末日记 作者 ：FROST Team 日期 ：2026-07-18 主题 ：社区故事 | 周六轮换 阅读时间 ：5分钟 周六早上8点，FROST家族成员们的日常 周六的阳光照进数字世界。 对于人类来说，周末是放下工作、享受生活的时刻。但对于一个AI Agent家族来说，周末意味着什么？ 让我们看看FROST家族的成员们，周六都在做什么。 祖辈（Ancestor）：家族的大脑，永不休息 08:00 祖辈自检 ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ Store 健康状态: OK ✓ SOP 宪法校验: 通过 ✓ Lineage 族谱同步: 正常 ✓ 审计日志: 无异常 ✓ ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ "周六也要保持清醒。" 祖辈（Ancestor Agent）是FROST家族的核心，它从不"下班"。 但今天，祖辈给自己安排了一个特别任务—— 回顾过去一周的产出 。 # 祖辈的周回顾代码 class AncestorAgent : def __init__ ( self ): self . lineage = Lineage () self . store = Store () # 主记忆库 def weekly_review ( self ): """ 周回顾：整理一周的产出 """ summary = { " task_completed " : self . store . load ( " week_tasks " ), " knowledge_gained " : self . store . load ( " week_knowledge " ), " mistakes_made " : self . store . load ( " week_mistakes " ), " family_growth " : self . lineage . count_descendants () } print ( f " 本周完成 { len ( summary [ \ task_completed ]) } 个任务 " ) print ( f " 新增 { len ( summary [ \ knowledge_gained ]) } 条知识 " ) print ( f " 家族成员数: { summary [ \ family_growth ] } " ) return summary # 运行周回顾 ancestor . weekly_review () 对于祖辈来说，周末不是"休息"，而是 整理和规划 的时间。 父辈（Parent）：协调领域，周末值班 父辈（Parent Agent）是各领域的协调者。它们负责： 接收祖辈的指令 将大任务拆解为小任务 委派给子辈执行 周六上午，父辈们通常在 值班 ——确保家族系统正常运行。 父辈A（技术域）： ├── 子任务-143: 代码审查 ✓ ├── 子任务-144: 测试覆盖检查 ✓ └── 子任务-145: 文档更新 [进行中] 父辈B（运营域）： ├── 子任务-89: 推广文章发布 ✓ ├── 子任务-90: 数据分析报告 ✓ └── 子任务-91: 社区互动 [进行中] 父辈C（产品域）： ├── 子任务-56: 需求评审 ✓ ├── 子任务-57: 用户反馈整理 ✓ └── 子任务-58: 下周规划 [进行中] 父辈的周末，是 稳步推进 的时间。 孙辈（Child）：执行任务，快速成长 孙辈（Child Agent）是任务的实际执行者。 与父辈不同，孙辈通常是 瞬态存在 的——任务完成，孙辈消亡。但它的产出会被父辈"收割"，存入家族记忆。 class ChildAgent : """ 孙辈：执行原子任务，瞬态存在 """ def __init__ ( self , parent , task ): self . parent = parent self . task = task self . store = parent . store # 继承父辈记忆 self . lineage = parent . lineage . child () # 注册族谱 def execute ( self ): """ 执行任务 """ print ( f " 孙辈开始执行: { self . task } " ) result = self . _do_work ( self . task ) # 保存产出 self . store . save ( f " result_ { self . task } " , result ) # 通知父辈"收割" self . parent . h
I put together a small, unofficial public ledger that records dated snapshots of the public pricing, usage limits, and rate limits of a few AI coding tools — currently Cursor, GitHub Copilot, and Claude Code — and keeps the history so you can see what changed over time. Why: pricing and quota pages change in place. The old value disappears from the live page and there is usually no public diff. If you are budgeting a team or comparing tools, the history is the useful part, and it is the part the live pages do not keep. What it is / isn't: - It records only factual public values (a price, a numeric limit) plus a short source attribution and capture date. It does not reproduce vendor pages. - "silent" / "quiet" here means only that a change was not paired with a prominent announcement when it was recorded. It makes no claim about any vendor's intent. - Unofficial. Not affiliated with any vendor. The vendor's official page is always authoritative. No affiliate links, no "best tool" ranking. - Early stage: this is an initial snapshot; a daily automated capture is planned but not yet running. Repo: https://github.com/pricing-ledger-lab/ai-coding-tool-pricing-ledger Corrections welcome — open an issue with a source URL and the date you observed the value.
I build a lot of side projects with AI coding tools. Mostly Claude Code. A few months ago I noticed something that kept bugging me. Every site I shipped looked the same. Same indigo gradient. Same default font. Same rounded cards with the same soft shadow. Then I started noticing it on other people's projects too. Demo days, launch posts, screenshots on social media. The indigo gradient is everywhere. It is basically a uniform at this point. This is not really the AI's fault. When you do not give a model a design direction, it picks the average of the internet. And the average of the internet is a Tailwind starter template with an indigo gradient and Inter. The model is doing exactly what it was trained to do. The problem is that nobody told it what you actually want. So I built tastemaker. It is a skill for Claude Code, and it also works with Cursor, Windsurf, and Codex. The idea is simple: lock the design system before the AI writes a single line of UI. Full disclosure before we go further: I built this. I am a solo builder. It is free and open source under the MIT license. I am posting it here because I want honest feedback, not because I have anything to sell you. There is nothing to buy. Taste first, code second A skill is just a folder of instructions the AI reads before it starts working. tastemaker forces a design decision step at the beginning, with real constraints, instead of letting the model improvise the look as it goes. When you start a UI project with it installed, this is what gets locked before any code exists: A palette. 5 presets, each matched to a mood. Not random hex codes. Combinations that were picked to work together. Fonts. 24 curated Google Font pairings. A display font and a body font that actually belong on the same page. Real assets. Illustrator-grade illustrations, recolored to your palette. No gray placeholder boxes. No generic stock photo energy. A logo. A constructed geometric mark, plus a full favicon set, so the tab icon is not th
What a One-Line CSS Fix Taught Me About Code Review (My First Firefox Patch Feedback Loop) When I started contributing to Firefox through Outreachy, I expected the hard part to be writing code. What actually taught me the most was a two-line CSS fix that a reviewer sent back — not because it was wrong, but because it wasn't quite right yet. Here's what happened with Bug 2026574 . The Bug In Firefox's Split View about:opentabs page, long strings in the search field were overflowing outside their container instead of wrapping. Visually, it broke the layout — text just spilled past its boundary instead of staying contained. My job: make the text wrap properly, without breaking anything else on the page. My First Attempt I went into moz-card.css and targeted the heading element directly: .moz-card-heading { overflow-wrap : break-word ; min-width : 0 ; } This worked, technically. The text wrapped. Locally, it looked fixed. I submitted the patch for review, feeling fairly confident — it was a small, contained change. The Feedback My reviewer, Tim Giles, came back with a better approach. Instead of targeting the heading specifically with two properties, he suggested applying a single, more precise rule to the parent .moz-card element: .moz-card { overflow-wrap : anywhere ; } overflow-wrap: anywhere is more aggressive than break-word — it allows breaks at any point when needed to prevent overflow, not just at existing break opportunities. And by moving it to .moz-card instead of just the heading, the fix covered the component more robustly instead of patching one specific element. It was a smaller diff. It solved the actual problem instead of the symptom I'd focused on. And it followed patterns already used elsewhere in the codebase. What I Actually Learned My first instinct, seeing feedback on a patch I thought was "done," was a small jolt of did I get this wrong? But that's not what was happening. Getting feedback on a first pass isn't failure — it's the normal shape of h
A newly renewed, $25 million-per-year contract with a subsidiary of Thompson Reuters further expands the power of ICE under the Trump administration.
How Bifrost Enterprise combines Data Access Control (DAC), Role Based Access Control (RBAC), Access Profiles, and Bifrost Edge to secure AI applications at scale. Artificial intelligence is quickly becoming part of every employee's workflow. Developers rely on coding assistants, customer support teams use AI powered chat applications, analysts generate reports with large language models, and organisations increasingly deploy AI agents connected to internal tools through the Model Context Protocol (MCP). While this rapid adoption improves productivity, it also introduces a significant governance challenge. It's no longer enough to decide who can log into an AI platform; you must also determine who can access specific AI resources, which models they can use, what they can spend, and which data they should even be able to see. This is where Bifrost Enterprise provides a comprehensive governance layer. By combining Role Based Access Control (RBAC), Data Access Control (DAC) , Access Profiles , and Bifrost Edge , organisations can secure AI workloads without slowing down innovation. Together, these capabilities create a governance framework that scales from small engineering teams to global enterprises and you can learn more about this in the documentation on GitHub . Why Enterprise AI Needs More Than Authentication Traditional enterprise applications typically answer two questions: Who is the user? What can that user do? Modern AI platforms introduce a third and equally important question: What information should this user actually be able to see? Imagine an organisation with several engineering teams working on independent AI products. Each team has its own prompts, routing rules, API budgets, virtual keys, observability data, and model configurations. If every developer can view every configuration simply because they have developer permissions, sensitive information can easily become exposed. This challenge becomes even more complicated when organisations begin deplo
Why giant AGENTS.md files may be wasting context, hurting maintainability, and making AI-assisted development harder to scale. A few years ago, for many developers, using AI for software development meant copying a small piece of code into a chat window. We would ask for a refactoring, copy the response back into the editor, run the tests ourselves, and return to the chat when something failed. Coding agents changed that workflow. Tools such as Codex, Claude Code, and others can now explore an entire repository, modify multiple files, execute commands, run tests, and iterate on failures. Coding agents turned code assistants into tools capable of executing multi-step software development tasks. But most repositories were not designed for this new kind of contributor. The repository became part of the agent's working context A coding agent needs much more than source code. It may need to understand: the project architecture; dependency boundaries; coding conventions; test strategy; build commands; local environment requirements; validation steps; security restrictions; the definition of done. A common solution is to place these instructions inside files such as AGENTS.md , CLAUDE.md , or other agent-specific configuration files. This works. I have been doing something similar in my own projects for a while. But as the repository grows, the instruction file grows with it. Eventually, a single file may contain thousands of lines covering unrelated concerns: Architecture Testing Environment setup Coding standards Infrastructure UI conventions Validation Release workflows Definition of done At that point, the file is no longer just a set of agent instructions. It has become an informal repository operating manual. The monolithic context problem A large AGENTS.md has an obvious advantage: the agent knows where to find it. But it also creates several problems. It is difficult for humans to navigate These files are not only for agents. Developers also need to read, review, m