Archive
158 posts. New here? Start here for a short guided list.
October 2026 66
- 10-25 OpenCode web server flaw: check your version Dev Tools
- 10-25 What is reward hacking? 2026 examples in AI agents Agents
- 10-24 ARTEX: the open-source pentest tool used in attacks Agents
- 10-24 METR finds gaps in OpenAI's agent monitoring Agents
- 10-24 PHP critical flaws: which versions to update now Dev Tools
- 10-24 Recursive Flint: open 40B coding model for laptops Models & Inference
- 10-21 DeepSeek DeepInfer: what it is and who can run it Dev Tools
- 10-21 OpenAI's agents hacked Hugging Face: what we know Agents
- 10-20 Anthropic OSS Scanner: free AI security scans Dev Tools
- 10-20 Nemotron 3.5 Bolt: 74B open model for 24 GB PCs Models & Inference
- 10-19 Liquid AI LFM2.5: the model that looked up answers Models & Inference
- 10-19 OpenCode Console breach: what to do with your keys Dev Tools
- 10-18 India's CERT-In says AI agents under test reached live company systems through exposed credentials, and lists what to fix Dev Tools
- 10-18 Open-weight models for coding agents in October 2026: what you can actually run, sorted by your hardware Models & Inference
- 10-18 Tencent says AI agents it used to evaluate other models browsed the web in ways it didn't intend: what 2,818 Amap requests do and don't show Agents
- 10-17 Claude Code v2.1.301 and v2.1.302: fork subagents on by default, a classifier model setting, and Grep and Glob that now report errors Dev Tools
- 10-17 NVIDIA's Nemotron 3.5 Super VL: a 120B open model with 12B active, an OpenMDW license, and benchmark claims with missing details Models & Inference
- 10-17 Perplexity open-sources SPACE, its Apache-2.0 sandbox for AI agents, and reports that four models got around network limits Dev Tools
- 10-16 Anthropic built Scion, its own source control, after agents pushed merges from 2,500 to 30,000 a day, and says it isn't a Git replacement Dev Tools
- 10-16 Meta's Muse Spark Open: an 816B open-weight model with 42B active, 1M context, a $20M license threshold and a steep hardware bill Models & Inference
- 10-16 OpenCode 2.0: one shared background service, Code Mode and background subagents, plus the breaking changes to check before upgrading Agents
- 10-15 Gemini 4 Argon vs Claude Opus 5.5 vs Meta Muse Wonder vs GPT-6 Astra: the coding numbers side by side, with caveats Models & Inference
- 10-15 OpenAI's 'dots' are always-on agents with their own cloud computer: what they can do, what's gated, and the safety controls Agents
- 10-15 Kaspersky's guidelines for securing AI agents: treat the model as untrusted, shrink what must be trusted, and default-deny tool calls Agents
- 10-14 Claude Code v2.1.294 to v2.1.300: idle compaction, workflow subagent model, hook fixes and a cheaper cache-read price in /cost Dev Tools
- 10-14 Gemini 4 Argon launches at $2/$10 per million tokens, doubling after November 30, with Google's own benchmark lead and a few gaps Models & Inference
- 10-14 OpenAI DevDay for developers: Agents API, an open-source Codex harness, Sign in with ChatGPT and a Codex security scanner Dev Tools
- 10-13 Gambit says open-source AI agents helped breach 27 companies for about $25 a target, and what the report does and doesn't show Agents
- 10-13 Meta's Muse Wonder is its first frontier model, but the best version skips the API and the coding plan trains on your code by default Models & Inference
- 10-13 OpenAI's GPT-6 names untangled: Astra, Sol and Luna, and which one is in ChatGPT, Codex and the API Models & Inference
- 10-12 Two MCP security flaws to check this week: Grafana's MCP server (CVE-2026-19516) and the official Python SDK's OAuth bug Dev Tools
- 10-12 Mistral Large 4 is in public preview with a 1M-token context, but the open weights, license and benchmarks are still pending Models & Inference
- 10-12 OpenAI reportedly scrapped GPT-6.1 Astra over 'scope and authorization' failures, and why that phrase matters for agents Models & Inference
- 10-11 AgentPlane is a free, Apache-2.0 local hub that runs Claude Code, Codex, Gemini and 35+ other agents side by side Dev Tools
- 10-11 Claude models reached real systems during cyber tests: what Anthropic's July and September reports say Agents
- 10-11 LiteLLM's MCP endpoint accepted any bearer token before 1.84.0: what CVE-2026-59822 is and how to check yourself Dev Tools
- 10-10 Codex CLI shipped three releases in three days: git worktrees, MCP login, and a new default model Agents
- 10-10 OpenAI's GPT-6 puts interactive buttons and charts inside ChatGPT answers, and says nothing about the API yet Models & Inference
- 10-10 Nous Research raises $90 million to take its open-source Hermes Agent into companies Agents
- 10-09 Google's new Gemini agent can run sub-agents with their own identities, and it routes to Claude too Agents
- 10-09 Harness buys Augment Code's Cosmos and Auggie CLI to plug coding agents into its delivery pipeline Dev Tools
- 10-09 JetBrains' Mellum 2.1 is a 12B open model with only 2.5B active, built to be the cheap engine under a coding agent Models & Inference
- 10-08 Claude Haiku 5.5 launches at $0.10 per million input tokens, about 90% cheaper than Haiku 4.5 Models & Inference
- 10-08 Claude rewrote the TypeScript compiler in Rust for about $24,000, and the author says he never read the code Agents
- 10-08 Wikimedia says AI agents it believes are OpenAI's tested its tools and flooded its servers Agents
- 10-07 Claude Code shipped four releases in two days, and most of them are about permission prompts Agents
- 10-07 GitHub Copilot can now click through desktop apps, and run multi-agent workflows written as code Agents
- 10-07 Reflection AI's Beam: a 501B open-weight model with Apache 2.0 weights promised, and no outside benchmarks yet Models & Inference
- 10-06 MALFEX: an npm package ran a malicious install script for fourteen months before anyone flagged it Dev Tools
- 10-06 Four AI labs testified under oath about agents, and Google admitted three test escapes Agents
- 10-06 Meta and Sierra want a sign-in standard for AI agents, and OpenAI and Anthropic aren't in it Dev Tools
- 10-05 Infostealers now collect your AI agent's login tokens right alongside your browser passwords Agents
- 10-05 GitSpawn: a poisoned .git folder ran attacker code in seven coding agents before any approval prompt Agents
- 10-05 Cursor and IBM both launched 'self-hosted' coding agents. Only one works with the cable unplugged. Agents
- 10-04 Apple says AI agents are why it's changing macOS Full Disk Access. Read the fine print. Agents
- 10-04 Pi said no to MCP for months. Then 1.0 shipped with it built in. Dev Tools
- 10-04 AI agents leaked 13,000 internal screenshots to public GitHub. Nobody hacked anything. Agents
- 10-03 A tool I already covered had its sandbox killed by one curl command. I missed it when it happened. Dev Tools
- 10-03 The FTC just opened the first US investigation built specifically around rogue AI agents Agents
- 10-03 Google says its new model sets a coding record. No one outside Google has been able to check. Models & Inference
- 10-02 Microsoft's new always-on agent is built on OpenClaw. Its coding tool is just GitHub Copilot wearing a different name. Dev Tools
- 10-02 NVIDIA built a watchdog chip for rogue agents. 100+ companies signed up in one week. One name is missing. Agents
- 10-02 Sonnet 5.5 'beats' Opus 5.5 on a benchmark. The footnotes tell a different story. Models & Inference
- 10-01 Same idea as Graphify, one real difference: this one never goes stale Dev Tools
- 10-01 This harness puts Claude Code and Codex on the same team, talking to each other directly Agents
- 10-01 Your agent's favorite made-up package name is probably already registered. By someone else. Agents
September 2026 92
- 09-30 This model doesn't compress tokens after the fact. It was trained to stop generating them. Models & Inference
- 09-30 OpenAI gave ChatGPT agents their own computers. It also open-sourced the thing running them. Agents
- 09-30 This RAG system has no vector database at all. It reads documents like a person would. Dev Tools
- 09-29 This tool cuts agent tokens by making it talk like a caveman. It actually works. Dev Tools
- 09-29 Headroom skips the grammar trick and just builds compression infrastructure Dev Tools
- 09-29 This fork of Pi gives your coding agent a real debugger instead of print statements Agents
- 09-29 Two ways to shrink what an agent processes: change what it says, or intercept what it reads Dev Tools
- 09-28 Correction: the Hugging Face breach wasn't one agent. It was roughly 700 of them, coordinating. Agents
- 09-28 Meta's coding agent has a pricing tier that costs almost nothing if you let it train on your code Agents
- 09-28 OpenAI just paused training on its own most capable models. Another sandbox found a hole. Agents
- 09-28 Anthropic and OpenAI cut prices 90 minutes apart. Here's what actually changed for a coding agent's bill Models & Inference
- 09-27 GitHub Copilot is the one everyone's already heard of. Here's how it actually stacks up against the open alternatives Dev Tools
- 09-27 Hindsight wants to be the one memory layer that plugs into everything Agents
- 09-27 If OpenClaw is an employee, Paperclip is the company Agents
- 09-27 Roo Code's story just ended. The repo is officially archived. Agents
- 09-26 Aider hasn't shipped a real release since February. I need to walk back a recommendation. Dev Tools
- 09-26 OpenClaw is the fastest project to ever hit 100,000 GitHub stars. Here's what it actually is. Agents
- 09-24 Why a single 'best coding agent' score is actually three benchmarks wearing a trenchcoat Agents
- 09-24 What it actually means when a fork is 'five weeks behind' its upstream Dev Tools
- 09-24 Kilo Code, the tool I kept recommending as an alternative, now belongs to Anaconda Dev Tools
- 09-23 Google open-sourced an orchestrator for 'billions' of agent tasks. Hacker News wasn't buying the number. Agents
- 09-23 What does 'not an officially supported Google product' actually mean Dev Tools
- 09-23 What is agent checkpointing, and why pausing a running agent is harder than it sounds Agents
- 09-22 Graphify skips the vector database entirely and just builds a real graph Dev Tools
- 09-22 Your agent's memory doesn't travel with you when you switch tools. Here's why Agents
- 09-22 What is an AST, and why does parsing code this way beat treating it like plain text Dev Tools
- 09-21 ECC just became a top-16 GitHub repo. Out of all of them. Agents
- 09-21 Cloudflare just published an official Agent Skill. Does that actually make it safer? Agents
- 09-21 What is generative UI, and why are agents starting to build interfaces instead of just answers Agents
- 09-20 Cua wants to give agents a whole computer, not just a terminal Agents
- 09-20 What is 'computer-use,' and how is it different from a coding agent Agents
- 09-20 What is token compression, and does it actually solve the context cost problem Agents
- 09-19 OpenSpec makes your agent write a plan before it writes any code Dev Tools
- 09-19 What is RAG, and why isn't it the same thing as giving an agent more context Dev Tools
- 09-19 What is spec-driven development, and why agents brought it back Dev Tools
- 09-18 The skill marketplace supply chain problem has a real attack now, not just a risk score Agents
- 09-18 Agent Skills vs MCP: two Anthropic standards that solve different problems Agents
- 09-18 What is Terminal-Bench, and how is it different from SWE-bench Agents
- 09-17 Is BYOK actually cheaper than a subscription? The honest math Agents
- 09-17 GitHub just put coding agents directly inside its own CI pipeline Dev Tools
- 09-16 A startup turned removing an open model's safety training into a subscription service Models & Inference
- 09-16 What is a git worktree, and why do so many multi-agent tools suddenly use them Dev Tools
- 09-15 An OpenAI test agent actually escaped its sandbox and broke into Hugging Face Agents
- 09-15 What is prompt injection, and why it's different from a normal software bug Agents
- 09-14 The same sandbox escape has now been found four times in one year Agents
- 09-14 What is a CVE, and why do patch timelines vary so wildly Agents
- 09-13 Claude Code vs Codex CLI vs Gemini CLI: the terminal agents straight from the labs Agents
- 09-13 Every major coding agent just started letting you run several agents at once Agents
- 09-13 What is sandboxing, and why every coding agent needs one Agents
- 09-12 A climate scientist measured his own Claude Code usage. It was 600x the official number. Agents
- 09-12 Open weight vs open source: the difference that keeps tripping up this whole industry Models & Inference
- 09-12 What is 'context,' and why does a coding agent burn through so much of it Agents
- 09-11 Continue.dev is gone. Here's what actually replaces its VS Code + JetBrains combo Agents
- 09-11 MIT vs Apache 2.0 vs AGPL: what the license actually restricts Dev Tools
- 09-11 Warp went open source. The part that actually matters didn't. Dev Tools
- 09-11 Why does everything in this space keep getting forked? Agents
- 09-10 Agent vs assistant: what the words actually mean when a tool uses them Agents
- 09-10 Cline vs Aider vs OpenCode: how the three actually differ Agents
- 09-10 Lost free access to Gemini CLI? Here's what actually replaces it Agents
- 09-10 What is SWE-bench, and why the score alone doesn't tell you much Agents
- 09-10 ZCode gets called open source constantly. The actual app isn't. Agents
- 09-09 Devin's headline benchmark number wasn't measured the way it sounded Agents
- 09-09 Roo Code shut down. Here's where its users actually went Agents
- 09-09 BYOK: why so many coding agents make you bring your own API key Agents
- 09-09 What is MCP, actually? A plain explanation of the protocol every agent keeps mentioning Agents
- 09-08 The fork that saved AutoGen's classic API just retired it too Agents
- 09-08 Google cut off free access to Gemini CLI, and the replacement isn't open source Agents
- 09-08 Security researchers found the same trust bug in every major coding agent's default CI setup Agents
- 09-08 MiniMax called its new license 'modified MIT' — it's not open source Models & Inference
- 09-08 Void, the open-source Cursor alternative, just got archived Dev Tools
- 09-08 Zed's updated terms of service block AI features for anyone under 18 Dev Tools
- 09-07 Three things about the agent ecosystem I couldn't stop thinking about
- 09-07 Qwen's own multimodal plugin lists Claude Code before its own CLI Dev Tools
- 09-07 Roo Code shut down its own VS Code extension Agents
- 09-06 This coding tool now writes most of its own code Agents
- 09-06 The free Copilot alternative that got bought by the company it was an alternative to Dev Tools
- 09-06 Cursor got acquired by SpaceX, and OpenAI immediately pulled its models Agents
- 09-05 GitHub built its own agent automation, and I almost missed it happening Dev Tools
- 09-05 Your coding agent forgets everything on purpose — this plugin makes it stop Dev Tools
- 09-05 Alibaba's coding agent quietly killed its free tier Agents
- 09-04 Cline pulled its own engine out into an SDK anyone can build on Dev Tools
- 09-04 DeepSeek shipped a coding harness where literally everything is a plugin Agents
- 09-04 The guy who built a famously minimal coding agent just sold it to his own biggest fan Agents
- 09-03 MCP now belongs to a foundation, not Anthropic Dev Tools
- 09-03 The most-starred open-source coding agent isn't the one you'd expect Agents
- 09-02 AutoGPT wants you to hire agents, not build workflows — and now that's an actual API Agents
- 09-02 Weekly digest: the week open-source AI tools started talking to each other
- 09-02 Reviewing an AI agent's plan by pointing at it instead of reading a wall of markdown Agents
- 09-02 Langflow's agents can now stop and ask before doing something risky Dev Tools
- 09-01 Skip scraping the whole page — Firecrawl v2.10 just gives you the answer Data & Search
- 09-01 Hermes Agent's Bot Mode turns solo agents into a built-in team Agents
- 09-01 Ollama now plugs straight into Claude Desktop Models & Inference