Reprints, Photocopies and PermissionsCustom Reprints: Custom Reprints (minimum order 1,000). Phone: 212-512-4170; Fax: 212-512-6243; Email: [email protected] Black and White Photocopies: Contact ...
Show HN: GitHub Dashboard aka GitBiasedHi all.I'm working on GitBiased platform. Basically the idea is to let you or your team (manager) know what is happening at the moment in your development life.I myself use it lot, for having PRs of all my GitHub projects in one place. Also you can enable weekly mail with all things you have done for the previous week.Permission-wise each widget requires its own permissions. At the moment we do not require any write permissions first of all and none of them allow to read the code of the repo(s).I'm writing this post because I've recently released iOs version of the app.https://gitbiased.com https://apps.apple.com/gb/app/gitbiased/id6782933062 GitBiased
ChatGPT Can Take Over Your iMessages. But Do You Really Want That?If you want to add the new OpenAI feature, go to the ChatGPT Plugins menu, select Apple Messages and install it. You’ll then see a permissions screen in ChatGPT indicating that the message history on ...
Show HN: WaveHouse – Supabase for ClickHouseWhile building an IoT telemetry solution, we ran into hurdles with Clickhouse. For one, you can't insert quickly AND durably into Clickhouse without setting up something like Kafka, which gets complicated for quick projects wanting to make use of Clickhouse's powerful features. Then, trying to actually query Clickhouse and show data in a UI required a whole backend API to handle auth and permissions.We figured that all these parts together – fast, durable ingest, row-level and column-level security and roles, and realtime streaming – were a lot of scaffolding to have to rebuild for every project we wanted to use Clickhouse in. So, we built them all into a single Go binary to be deployed alongside Clickhouse, to help lower Clickhouse's barrier to entry. We call it WaveHouse.Would love any feedback as we work on improving and adding more features to this OSS project!
Safer Smartphones: Key Settings To Check For Better SecurityPeriodically reviewing key settings and content can help users spot outdated permissions, unnecessary access and issues that may affect privacy, security or efficiency.
Show HN: Csift – the missing tool to sift your Claude Code sessionsI built csift to address all my complaints about Claude's handling of its own session files. Claude Code sessions are plain-text JSONL with everything, but they never document the structure anywhere. Every time, Claude has to craft an ad hoc tool to guess, and SKILL.md isn't enough to cover everything.This is what happened in the night that made it a real tool: Claude deleted project files, froze, and said sorry at midnight. I didn't Ralph loop it, and I asked it to recover from its own JSONL and continue. I thought it would be easy, until it stopped again in the middle of rewriting them, since they're scattered across subagents, with its 5hr limit burnt out. This wasn't the first time after I used --dangerously-skip-permissions. And I created the tool that could put the file back together byte-exact.Now csift recovers deleted files and plans, reconstructs turns, matches plans to sessions, does typed search over 25 record labels, extracts pasted images, and maps the subagent topology t
Show HN: Burla – Distributed computing framework for AI agentsHi HN! Would love any thoughts on this, we built Burla to enable anyone, even total beginners to scale Python to thousands of computers in their cloud with zero hassle. In a world of coding agents, this means something different than it used to, specifically that Burla requires almost no cloud permissions to get started. Anyone who has permission to boot a VM in their cloud can simply pip install burla and scale Python to 1000's of VM's. Burla uses your local aws cli credentials to boot VM's and the dashboard runs locally for you to monitor resources and jobs.All together this creates a UX where you can simply ask your agent to do something like "Vector embed all of Wikipedia" and it will do it far faster and with less compute than it would have using any cloud service.
Ask HN: How can I find trackers I did not place in my stuff?By sheer luck I found a Tile BLE tracker in a car I purchased recently. I assume it was put there by the previous owner in case of theft and forgotten. It nonetheless has been destroyed. Now I'd really like to know how you defend yourself against this sort of attack?I'm not looking to defend against a 3 letter agency or nation-state, but against any low tech stalker, ex, would be thief, or cop without access to Flock cameras type of adversary. Looking for something that would be able to find BLE trackers regardless of platform/company.I ran a scan using nRF but there is a lot of noise out there and isn't sustainable. Have you found trackers in your stuff? How do you defend against trackers? There are apps out there but it's hard to trust them with the required permissions. Open source would be great. I'm on Android and have "tracker protection" or whatever enabled.
Show HN: Goodfault: insurance for AI agents and robotsHi HN. I'm building GoodFault, insurance for companies whose AI agents screw up: wrongful refunds, hallucinated policies a court makes you honor (the Air Canada case), deleted prod databases,leaked data. Two things happened this year that made this a real market. ISO shipped exclusion endorsements (CG 40 47 family) that strip generative AI from standard general liability at renewal, and courts kept assigning agent mistakes to the deploying company, not the model vendor. The interesting technical problem is pricing. Model identity turns out to be almost useless as a rating variable: the same model is a rounding error wired to a read-only knowledge base and a catastrophe wired to refunds. So we price the authority envelope instead: unattended financial cap, reversibility, reach, and action rate, with hard underwriting gates (no kill-switch, no logging, no injection testing = uninsurable). This page is that rating logic, public: set your agent's permissions and see the worst-weekend loss
Poker and Liar's Dice over Bluetooth Low EnergyI remember being able to play poker and trivia with random strangers on cross country flights on the IFE screens. For some reason, they took that away, so I decided to make some myself. The first two games are here: Poker - * iOS: https://apps.apple.com/us/app/airplane-mode-poker/id6791247916 * Android: https://play.google.com/store/apps/details?id=de.testco.airplanemode.pokerLiar's Dice - * Android: https://play.google.com/store/apps/details?id=de.testco.airplanemode.liarsdice * [iOS still in review]BLE was the choice to have it work cross-platform and support multiple players. There is no Internet access and there is no analytics or app tracking. The game literally doesn't ask for the permissions to access the Internet.It costs $2.99 as a one time unlock to host and $0 to join a game.Thoughts on other games?
Ask HN: AI rollout for 100 employees, mostly non-technical – what worked?Looking to roll out AI tools to my corporate team of around 100 employees. We have some engineers, but most are non-technical people (ops, finance, marketing, and site-level managers).What I want to avoid:- everyone pasting company data into random consumer AI tools with zero visibility or control. - everyone making their own version of the same tool over and over. - lots of different server instances all over the place that IT doesn't know about.What I'm looking for:- Sanctioned AI access with some admin visibility (usage policies, data retention, permissions by team) - Ideas for shared infrastructure and how to collaborate on projects better. - Visibility for teams to see what other people have built. - Encourages adoption/collaboration (shared prompts, internal use cases) rather than locking things down so hard nobody uses itI really appreciate the HN community because a lot of you are way ahead of me in this process. If you have found any good tools or ways to use existing tools or
Cloudflare open-sources vibe-coding platform for people who aren’t codersTo minimize the exposure of company data, AI agents start out with no permissions to access or share resources and must request them through the Cloudflare OS platform. Server cod ...
Show HN: Watchfire, an open-source control room for AI coding agentsHey, built watchfire because I wanted a way to run multiple agentsnat the same time without having to click accept permissions. That evolved into a way to do it securely, which then evolved into a spec-based dev approach with tasks. Last addition was to add support to expose watchfire as an mcp server. Would love to get feedback.
Show HN: Keystroke – open sourcing our internal agents and automations platformHello HN,I’m Dallin, building Keystroke with my co-founder Blake. Keystroke is a recent pivot from Buster, our original YC W24 company.Keystroke lets you build agents and durable workflows as TypeScript in your own repo, then deploy them to a web application where your team can use them. Our repo (ELv2 licensed) is here: https://github.com/keystrokehq/keystroke.At Buster, we spent a lot of time building internal agents and automations. We used visual platforms like n8n and Zapier, as well as code-first tools like Trigger.dev, Mastra, and several of Vercel’s open frameworks. We genuinely liked many of them, but kept encountering two different problems:1. Visual platforms gave us integrations, credentials, hosting, and an app the rest of the team could use, but became difficult to test, review, and debug as our systems grew.2. Code-first tools gave us the flexibility of ordinary TypeScript and worked well with coding agents, but left us building the UI, runtime, permissions, credentials,
Show HN: Sixb – the operating layer for enterprise AIHi HN!We open sourced a framework to be able to model the operational layer of a business.1. The problem today:Almost every company uses an AI assistant: ChatGPT, Claude, or Gemini.But it remains an assistant.And the companies are giving more and more information, even if some are sensitive, because better context usually produces better answers.The model should not contain the company’s operations. It should be one component connected to them.2. Our thesisAI is making interface and glue code cheap.The difficult part is modeling the business context, rules, permissions, and workflows behind them.MCP solves access to tools, but it does not solve business context, permissions, or governance.You cannot get a useful AI implementation off the shelf. Every company operates differently.3. What you can do with Sixb- Connect your data: CRM, ERP, file systems, meeting transcripts... - Model your business: Customers, projects, equipment, quotes, contracts, and the relationships between them - Def
Show HN: Brainstorm – a local-first, AI-native OS for knowledge workHello,I've been building a local-first open-source knowledge management operating system for 3 months as an AI-native company and would love to hear genuine feedback from the Hacker News community. I've previously worked on a similar project for several years and tried to solve a problem that everything is glued together in common knowledge management apps and fixing one thing usually led to breaking other user experiences. What I was trying to build is disconnecting things and making apps first class citizens, responsible for one single piece of user workflow, like journalling or using whiteboards for example.As the app I was working on had heavy security measures to preserve data sovereignty, I decided that I want the security to be weaved into the system at the core, so all apps are small pieces of software, sandboxed and with their own permissions and capabilities. Furthermore, the system now is a marketplace of assets - apps, themes and agents next. Another major feature is that w
A 1989 Unix finger protocol prank at UW-MilwaukeeIn 1989, as a CS undergrad at UW-Milwaukee, I had access to a graduate Unix lab. One afternoon, the grad students discovered they could scan the network for users who had accidentally left their .plan files world-writable. For the uninitiated, the finger command let you see who was logged in, and it displayed their local .plan file. Because default security permissions were incredibly relaxed back then, you could sometimes write directly to someone else's file. The grad students wrote a script to append a snarky message mocking the users' permissions: "I AM FAR MOR THAN AN IGNORANT MORON..." Later that night, the lab cleared out. Left to my own devices, curiosity got the better of me. I ran a similar scan to see if any open files were left. I found one, and the user’s login name immediately caught my eye: "beaver". The temptation was simply too great. I opened their file and wrote the only logical sentence:"If you are reading this, then you are fingering the beaver!" After literally fa