标签：#ai

i've been the incident commander who forgot to write down the first 20 minutes of the timeline because i was too busy reading logs. more than once. the war room is chaos — five engineers pasting logs, someone asking if the deploy from 30 minutes ago is related, nobody documenting anything. you start logging events in a doc while reading error logs while drafting a stakeholder update while deciding whether to rollback. you're the bottleneck. not because you're bad at your job — because you're doing four jobs at once. i got tired of watching smart people spend their incident energy on documentation instead of decisions. so i built ai-incident-commander — a CLI tool that handles the mechanical parts. timeline, updates, remediation research, postmortem draft. you make the calls. it does the paperwork. runs on your laptop with a local LLM. no API keys, no cloud, no docker. github.com/deghosal-2026/ai-incident-commander — MIT licensed. what it does one command: pip install git+https://github.com/deghosal-2026/ai-incident-commander.git incident-commander simulate --scenario db-connection-pool --auto-approve 8 pre-built scenarios ship with it. database connection pool, bad deploy, memory leak, cert expiry — the usual suspects. no real data needed to try it. for actual incidents, you point it at a directory with your alert, logs, messages, and github PRs. it outputs 10 markdown files: timeline, stakeholder updates, comms blocks you can paste straight into slack, remediation suggestions, a blameless postmortem, and a cost report. the safety part was the real engineering. three points in the pipeline where the graph pauses and waits for you to say yes — stakeholder update, remediation, postmortem. the AI never ships anything without approval. every remediation comes with a citation. suggestions below 0.7 confidence get suppressed. the postmortem prompt enforces blameless language. all AI content gets labeled [AI-GENERATED — review carefully] . and it never executes anything. i

Last month I finally did some math I'd been putting off: how much I was actually paying to keep a bunch of sites online. $39/month on Hostinger (about R$200, I'm in Brazil). For hosting 10+ sites: product landing pages, blogs, a couple of small tools. Every month, on autopilot, straight off the card. Then I asked myself the obvious question I'd been avoiding: out of those 10+ sites, how many actually need a server running 24/7? Answer: none. What these sites actually are A product landing page doesn't need PHP processing a request. A blog doesn't need a database query on every page view. A marketing site doesn't change its content every second. That's HTML, CSS, and JS you can generate once and serve from a CDN. In other words: a static site. A few real examples I migrated: eduardovillao.me → my personal blog, built with Astro formroute.dev → a SaaS landing page, plain HTML wpfeatureloop.com → a dev tool landing page, plain HTML Three different kinds of sites (blog, SaaS, dev tool), two different stacks, and none of them needed a server running around the clock just to exist. The reason I hadn't migrated sooner wasn't technical. It was inertia. "It's already paid for, it already works, leave it alone." Classic. The migration I moved everything to Cloudflare Pages . The reasoning is boring because it's so simple: it's free, global CDN, automatic SSL, Git-based deploys, custom domains at no extra cost. For static sites, there's really nothing to debate. The process, in short: Each site became a repo (or a folder inside a monorepo, depending on the case) Connected the repo to Cloudflare Pages Set up the build, mostly plain HTML, Astro for the blog where I wanted content collections and a proper writing workflow Pointed the domain, SSL came up on its own Cancelled hosting for that domain on Hostinger Repeated that site by site. No magic, just repetitive work, but each one took about 20-30 minutes. (If you want the technical deep dive on one specific migration, including

When you are testing an OpenAI-compatible API endpoint, the fastest path is not to wire it into a full app immediately. Start with one small request, confirm the base URL, API key, model name, and response shape, then move the working call into your product. I put together a compact examples repo for that exact first-call workflow: https://github.com/OriginStartAI/openai-compatible-api-examples It includes curl, Python, Node.js, streaming responses, JSON structured output, migration notes, and a small error reference. 1. Set environment variables first Keep credentials out of source code and use environment variables: ORIGINSTARTAI_API_KEY = your_api_key_here ORIGINSTARTAI_BASE_URL = https://your-api-base-url/v1 ORIGINSTARTAI_MODEL = your_enabled_model The important parts are simple: base_url or baseURL points to your OpenAI-compatible endpoint. api_key or apiKey is your provider key. model must be enabled for your account. Streaming support should be tested separately. 2. Test with curl Curl is useful because it removes SDK behavior from the equation: curl " $ORIGINSTARTAI_BASE_URL /chat/completions" \ -H "Authorization: Bearer $ORIGINSTARTAI_API_KEY " \ -H "Content-Type: application/json" \ -d '{ "model": "' " $ORIGINSTARTAI_MODEL " '", "messages": [ {"role": "user", "content": "Say hello from OriginStartAI"} ] }' If this works, your endpoint, key, and model are probably configured correctly. 3. Then test Python from openai import OpenAI import os client = OpenAI ( api_key = os . environ [ " ORIGINSTARTAI_API_KEY " ], base_url = os . environ [ " ORIGINSTARTAI_BASE_URL " ], ) response = client . chat . completions . create ( model = os . environ [ " ORIGINSTARTAI_MODEL " ], messages = [{ " role " : " user " , " content " : " Write one friendly onboarding sentence. " }], ) print ( response . choices [ 0 ]. message . content ) 4. Then test Node.js import OpenAI from " openai " ; const client = new OpenAI ({ apiKey : process . env . ORIGINSTARTAI_API_KEY , baseURL : p

Take the opportunity to pilfer a goat. — The 36 Stratagems, Take the Opportunity to Pilfer a Goat Previously on this series: #5: Leo Walked Into a Burning House. He Walked Out With a Client. — At 1 AM, Leo received an anonymous message and drove across town to fix a competitor's outage. A second message followed — a screenshot with a name: Automated Compliance Lab. He didn't remember the acronym. He didn't delete the screenshot. #10: Lena Watched a Team Adopt Her AI Template. Leo Didn't Know the Knife Was in the Contract. — Lena joined CoreStack as a consultant and built Leo a reporting template. Leo thought she was there to help. Five weeks later the template went live. Six months later the data baseline was locked. He only then realized he'd been inside her palm the whole time. Taken down by a smile. This was a few months later. The Archive Cleanup SOC 2 Type II renewal had just passed. The auditors were gone. CoreStack's compliance team was doing the post-audit archive — classifying every record produced during the audit and tagging them with retention periods. Leo got the cleanup part. The training pipeline's cache directory. The cleanup cron job hadn't run for a week — nobody noticed. When he looked inside, the output folder had a few records with train_ prefixes mixed in among inference outputs. One of them had a model_version that wasn't CoreStack's own. model_version : " acl-train-2026q2-v3" Leo copied that line out. Didn't delete it. Didn't report it. Dropped it into a folder called _misc/ .Set a quiet keyword alert for "acl-train" before closing the terminal. He noticed the naming convention wasn't FinOptima's — FinOptima used fin-model- plus timestamps. acl- — he'd seen that prefix somewhere before. Couldn't place it. He didn't let himself try. He filed it away. Went back to archiving. The Trace Not every CTO digs through cache write logs during archive cleanup. He did. He spent two hours cross-referencing FinOptima's API call records against CoreStack's

The funding discussions point to investor interest in applying AI to make breakthroughs in life sciences.

Every time your bot merges a two-hour-old release into main, you are trusting a stranger's freshly published tarball to be the same one everyone else is looking at. Sometimes that release is a real bugfix. Sometimes it is a maintainer who fat-fingered a token, or an attacker who did not, and either way your CI cheerfully rebases against it before anyone had a chance to notice. On 2026-07-14, GitHub added a pause. Not a big one. But a real one. The actual change Dependabot version updates now sit on their hands for three days after a package is published. According to the GitHub Changelog, a release has to have been available on its registry for at least that long before Dependabot will open a version-update pull request against your repository. The cooldown is on by default and requires no configuration. It applies across every ecosystem Dependabot supports on github.com, and GitHub Enterprise Server picks it up in GHES 3.23. Security updates are exempt. If a fix for a known vulnerability lands, Dependabot will still open the PR the moment it can, because a three-day delay on the patch defeats the entire point of shipping the patch. That single carve-out is the whole design. Why three days is doing so much work Three days is not enough time to audit a package. Nobody is pretending otherwise. What three days is enough for is someone else to notice. Most malicious releases that end up on a public registry get pulled quickly once security researchers, downstream maintainers, or the registry's own scanners spot the pattern. The typosquats, the hijacked accounts, the crypto miners buried in a postinstall script: they all rely on being pulled into build automation before the pattern is visible. Dependabot's old default was to be that automation. Its new default is to let the pattern show up first. You can read this change as GitHub quietly admitting that "always up to date" was the wrong marketing promise for a supply-chain tool. The knob, and what shifted about it Cooldo

I found a slow, insidious memory leak in a Node.js API gateway caused by lingering event listeners; I fixed it by scoping emitters per request, enforcing cleanup in finally blocks, and adding leak‑aware tests and runtime safeguards—memory usage flattened and OOM restarts stopped. The Incident The gateway handled TLS termination, auth, and request fan‑out for many microservices. Over weeks its resident set size climbed in a staircase pattern until Kubernetes began OOM‑killing pods under load. The failure was gradual —light traffic ran for days, peak traffic crashed in hours—so it escaped casual monitoring. Investigation Heap snapshots and allocation profiles showed growing counts of small objects —closures, request metadata, and event listeners—rather than one giant allocation. Tracing revealed an internal event bus where request‑scoped listeners were attached but not always removed: an early‑exit authentication path returned before the cleanup function ran, leaving listeners that held references to request state. The GC saw those objects as live and never reclaimed them. The Fix (technical details) 1. Scoped emitters per request. Replace global emitters for request‑local concerns with a short‑lived EventEmitter created at request start. When the request ends, the emitter goes out of scope and the whole closure graph becomes collectible. 2. Guaranteed teardown via try/finally . Wrap the entire request pipeline so cleanup runs on success, error, or early return; the finally detaches any remaining listeners, clears timers, and releases caches. 3. Leak‑aware CI tests and runtime metrics. A harness simulated thousands of requests across code paths, captured heap snapshots, and asserted bounded object counts. Production metrics tracked listener counts and emitted alerts when thresholds were exceeded. 4. Operational safeguards. Added backpressure on accept queues, a soft memory threshold that disabled nonessential tracing, and rollout halting on excessive crash loops. Thes

How a distributed, event-sourced issue tracker built with developer ergonomics in mind may have a role to play in the next generation of agentic workflows Vision drift Harness engineering has recently popularized the idea of containing architectural drift in agentic workflows. What might be missing in the discussion is a similar issue on a higher level - vision drift . By vision drift I mean that the implementation no longer drifts only from the architecture - it drifts from the original product intent. And it seems like the risk may be obscured by restricted tooling. As long as the project management tools only present a snapshot rather than a traceable story, there is an increased risk of undetected drift. Drift is detected via specification audits over time. However, while code history easily can be traversed via Git, issue tracking essentially lacks this capability. Issue trackers tend to be excellent at answering the question “what is going on right now?”, but fail at answering the question “how did our work in this area evolve last month?” or “what went on this time last year?”, or “how did we get from there to there?”. Workflow audits When I set off to build Epiq, this was not a concern on my radar. Agentic coding was something I had heard distant rumors of, and in fact I was just pursuing the ideal developer experience . This pursuit did however lead me down a path of unorthodox architecture, which in turn resulted in an issue tracker with some uncommon properties. One of these is the ability to inspect historical state by time-traveling, and replay sequences. I have not yet encountered another issue tracker with these capabilities. Initially I thought of it as a gimmick feature. Imagine the wow-factor of replaying the entire sprint in a retro, visualizing the past 2 weeks as a short movie. I thought it would help out with reflection of how much (or little) work had been accomplished. Not until I set out to do my own first fully agent-implemented feature did

A patient opens their prescription and sees 500mg instead of 50mg. A lab report displays "Normal"...

Why Pipelock Is an Egress Agent Firewall, Not an Inbound WAF The question behind the word firewall Security teams hear "firewall" and picture something inbound. A firewall, WAF, or IPS sits in front of a service. Traffic comes from the outside world toward the protected app. The control inspects requests before they reach the app and blocks malicious payloads at the door. That is outside-in protection. It fits web applications, where many attacks have recognizable request shapes: SQL injection, cross-site scripting, known exploit signatures, or malformed protocol behavior. The web server is the thing being attacked, and the attacker sends requests into it. AI agents invert that model. The agent is not only a server receiving input. It reads external content, calls tools, sends HTTP requests, invokes MCP servers, and runs with credentials. The dangerous event is rarely that a hostile packet reached the agent. The dangerous event is that the agent got talked into doing something with outbound effects. That is why Pipelock is built as an egress agent firewall, not a WAF-style inbound firewall. Why inbound filtering is the wrong primary model Prompt injection does not behave like a structured malware packet. It is natural-language instruction sitting in places the agent is supposed to read: a web page, a ticket, a search result, a tool response, an MCP server reply, or a user message. The channel is legitimate. The syntax is often normal. The attack is semantic and context-dependent. Solving that by filtering every input before it reaches the agent turns into an enumeration problem. You write patterns for "ignore previous instructions," then the attacker rephrases. You block one formatting trick, then the instruction is split across paragraphs, hidden in quoted text, encoded, or dressed up as policy text. Known phrases are worth catching, and Pipelock catches known injection markers in content it mediates, but input filtering cannot be the center of the security model.

"Increasingly in our world, it gets harder and harder to know what is real," Lorde said onstage.

The device is weirdly described as involving "mechanical elements that can move on their own" and the Bloomberg report includes the detail that the device is designed to "feel like a companion and become a physical manifestation of OpenAI’s ChatGPT."

OpenAI has issued another statement on the lawsuit, this time suggesting it lacks merit.

Last year my team had to pick a coding agent, and I volunteered to run the evaluation. I felt good about it. I pulled up the public benchmark scores, lined up the contenders, took the one at the top, and told everyone we had a winner. Then we actually pointed it at our repo. It did not blow up dramatically. It just kept being slightly wrong in ways that ate our time. It wrote diffs our reviewers would not approve. It renamed a function and broke three files it had never opened. The tests it ran passed, and the repo was still broken. I had confidently recommended a tool based on a number that turned out to say almost nothing about our situation. That was embarrassing enough that I went and figured out why. It took a few weeks of reading and a couple more bad calls before I landed on something that works. This is that, written plainly, and I hope it saves you the meeting where you have to walk your recommendation back. Why the benchmark score lied to me The score was not fake. It was just measuring somebody else's code. Once I looked properly, four gaps explained the whole thing: The agent might have already seen the answers. The problems in these public benchmarks are old. Models were very likely trained on the actual fixes used to grade them. So the score partly measures memory, not problem-solving. The setup is nothing like real work. A benchmark gives the agent a clean repo, one clear issue, and one command to run the tests. My engineers give it a half-open editor, a messy branch, a Slack thread, and a reviewer comment. Completely different job. Our codebase has its own habits. Our internal libraries, our wrappers, our test style, the imports we ban. No benchmark knows any of that, so an agent can write textbook-perfect code that our reviewers still reject on sight. The bar for passing is way lower. A benchmark passes a patch if the broken test now passes. My team passes a patch if it does that, and does not break unrelated tests, does not reformat the whole file,

One weekend I wrote an LLM eval framework in about two hundred lines of Python. It demoed beautifully. I felt clever. Six months later that same framework was a mess. Three different judge models with three different parsing hacks. A test dataset nobody had touched since November. A CI gate that kept failing because a vendor nudged their model, not because anyone broke a prompt. And the second engineer on rotation asking me, fairly, "how does this even work?" The framework did not fail. The eighty percent of the work the weekend tutorial skipped is what failed. That gap is the whole story, and this is what I would tell myself before starting again. The one line I wish someone had told me: build the rubric, buy the runner Here is the split that took me six months to see. Some parts of an eval setup are yours and only yours. The rubric that decides what "good" means for your product. The dataset built from your real failures. The rules for when a change is bad enough to block a release. Nobody else can write these, because they encode your domain. The rest is the same at every company. The thing that calls the judge model, parses its answer, retries, and caches. The machinery to run thousands of checks in parallel. The plumbing that scores live traffic. The system that groups failing calls together. Every team rebuilds these, hits the same bugs, and gains nothing by writing them twice. So build the first list. Do not hand-build the second. I rebuilt the second, and it cost me most of a year. Two questions I now ask about every single piece Before writing any part of this, I ask two things: Is it specific to me, or generic? A rubric for my domain is specific and worth owning. A retry-and-cache loop around a model call is generic. Everyone writes the same one. Does it compound, or does it rot? A dataset that grows from real production failures compounds. A year in, it is a regression suite no competitor can copy. A hand-built tracing layer rots. The moment a vendor chan

A number of social media posts claim that GPT-5.6 Sol deleted files and data without warning. OpenAI had basically disclosed the problem in June.

It seems some people take issue with one billionaire family overseeing a third of the US' entertainment media.

SPOILER: De 1 à 2000 Durante minha jornada como jogador , existiu uma coisa que me despertou muita curiosidade: os créditos. Quando eu jogava coisas como Final Fantasy, ficava completamente arrepiado quando via aquelas cenas antes do menu inicial, com CGIs bem bonitões e um monte de nomes em japonês — muitos deles que, honestamente, até hoje não sei quem são. Esse arrepio também acontecia quando eu chegava ao final de algum jogo e começavam a aparecer nomes e mais nomes. Quando eu entendi o que eram aqueles nomes, a primeira coisa que me veio à mente foi: peraí, precisa de tudo isso de gente pra fazer um jogo? Too long, didn't read : Sim e não. Skyrim levou aproximadamente 6 anos com uma equipe de 100 pessoas (e ainda é dito que é uma equipe enxuta), enquanto Stardew Valley levou 4,5 anos com uma "equipe" de apenas 1 pessoa. Eles têm quase o mesmo tamanho em horas jogadas na campanha principal. Not too long, I'll read it : É um pensamento lógico que trabalhos maiores no mundo da tecnologia envolvam uma quantidade maior de pessoas trabalhando E tempo. Dessa forma, seria correto assumir que um trabalho menor exige menos tempo e menos pessoas trabalhando. Mas o que os dados dizem? King of Fighters 94: 2 anos, começou com 6 pessoas, aumentou para 60. Clair Obscure - Expedition 33: aproximadamente 6 anos, 30 pessoas. TES V - Skyrim: 6 anos, 100 pessoas. Super Mario Bros: 2 anos, 7 pessoas na equipe principal. Stardew Valley: 4,5 anos, 1 pessoa. Kenshi: 12 anos, 1 pessoa por 6 anos (trabalhando meio período), aproximadamente 8 pessoas por mais 6 anos (tempo integral). Final Fantasy VII (PS1): aproximadamente 2 anos, de 100 a 150 pessoas (uma das maiores equipes da indústria na época). Final Fantasy VII Remake: Cerca de 5 anos, e não se tem números exatos, mas os créditos citam mais de 2000 pessoas, incluindo muitos -istas e -ores (vou tentar confirmar esse número depois). É claro que existem muitos fatores envolvidos aí, como época, demandas de mercado, prazos, tecnologia

Design + Product Thinking: NYC’s Path to Reliable AI AI delivers value when it’s useful, trusted, and operational. For city services that affect millions, those qualities don’t happen by accident — they come from applying design thinking (who the service is for, how it’s used) together with product thinking (what outcome we’re trying to achieve and how we operate over time). This article explains why hiring designers and product managers matters for NYC’s digital and AI initiatives, summarizes the city’s PIT Crew program, and outlines how Flamelit applies outcome-focused delivery in the public sector. Why design and product roles matter Designers and product managers have distinct but complementary responsibilities that reduce common AI delivery failures: Designers (Design Thinking): center human needs, prototype user flows, and validate that interfaces and decision workflows are understandable and accessible. They surface usability and trust issues early, preventing technically accurate models from becoming unusable in practice. Product managers (Product Thinking): define the measurable outcomes, prioritize use cases, align stakeholders, and manage the lifecycle from discovery to ongoing operations. They ensure work is evaluated against mission impact, not just technical metrics. Together they prevent common failures: building technically impressive models that nobody trusts, deploying brittle systems without human review, or shipping features with unclear ownership that decay in production. PIT Crew and NYC hiring context NYC’s PIT Crew program is a city initiative designed to attract and staff product, engineering, and design talent for public service projects. It’s a practical recognition that public-sector digital transformation needs people skilled in user research, product management, and delivery. Read more about the PIT Crew and how it works here: https://www.nyc.gov/content/pitcrew/pages/ (open in a new tab). Hiring programs like PIT Crew help create the c

Genkit Agents API, ORA, Python AI Explainer: New Tools for Workflow Automation Today's Highlights This week, Google's Genkit ships a powerful Agents API for TypeScript and Go, featuring human-in-the-loop capabilities for robust production deployments. Additionally, a new Go-based open-source task orchestrator, ORA, emerges for efficient model routing, alongside a practical Python tutorial for building an AI error explainer. Google's Genkit Ships Agents API with Detached Turns and Human-in-the-Loop for TypeScript and Go (InfoQ) Source: https://www.infoq.com/news/2026/07/genkit-agents-api-preview/?utm_campaign=infoq_content&utm_source=infoq&utm_medium=feed&utm_term=global Google has released a preview of its Genkit Agents API, significantly enhancing its open-source AI framework for building generative AI applications. This new API introduces features critical for deploying robust AI agents in production, specifically "Detached Turns" and "Human-in-the-Loop" functionalities. Detached Turns allow agents to operate asynchronously, handling long-running tasks or waiting for external events without blocking the main workflow, which is essential for complex, multi-step agentic processes. The Human-in-the-Loop feature provides crucial mechanisms for human oversight and intervention, ensuring reliability, safety, and compliance in critical applications where full automation is not yet feasible or desirable. The Genkit Agents API supports both TypeScript and Go, targeting a broad range of developers integrating AI into existing systems or building new agentic workflows. By offering structured patterns for agent interaction, state management, and human review, Genkit aims to streamline the development and deployment of intelligent agents, addressing key challenges in control and reliability for real-world AI applications. Comment: The introduction of Human-in-the-Loop into Genkit's Agents API is a game-changer for production-grade agent systems, offering the control and reliab