sf house sept 26
2 by jadenfix123 | 0 comments on Hacker News.
hey guys! I'm a 24 yo SWE @ scale and have previously worked at another YC startup. I'm passionate about building products, infra and eventually starting a company, it'd be great to live with other ambitious builders and founders. I'm looking for housing starting September 1st. If you know of a spot or are looking for a roommate with a similar mindset, lmk. my li is jadenfix
Hack Nux
Watch the number of websites being hacked today, one by one on a page, increasing in real time.
New ask Hacker News story: Cadreen – memory, governance, self-healing, and execution as one system
Cadreen – memory, governance, self-healing, and execution as one system
3 by ope_john | 0 comments on Hacker News.
Hi HN — I’m launching the research preview for Cadreen. Cadreen is infrastructure for intelligent workflows — software that can remember, decide, act, ask permission, and leave an audit trail. It includes APIs, SDKs, CLI tooling, memory, governance, tool execution, audit trails, and chat access. The idea is that useful AI systems need more than a model call. They need orchestration: what to remember, when to act, when to ask permission, what tools to use, and how to leave an audit trail. Available now: - APIs & docs - TypeScript, Python, and Go SDKs - CLI - Cadreen usable as a model/provider in chat interfaces (managed model selection with BYOK coming soon) - Coming next: “Accomplish Anything,” a chat interface for non-developers to use the same infrastructure directly. Feedback welcome — especially from people building agents, copilots, or AI workflow systems. https://ift.tt/xGTdvkC
3 by ope_john | 0 comments on Hacker News.
Hi HN — I’m launching the research preview for Cadreen. Cadreen is infrastructure for intelligent workflows — software that can remember, decide, act, ask permission, and leave an audit trail. It includes APIs, SDKs, CLI tooling, memory, governance, tool execution, audit trails, and chat access. The idea is that useful AI systems need more than a model call. They need orchestration: what to remember, when to act, when to ask permission, what tools to use, and how to leave an audit trail. Available now: - APIs & docs - TypeScript, Python, and Go SDKs - CLI - Cadreen usable as a model/provider in chat interfaces (managed model selection with BYOK coming soon) - Coming next: “Accomplish Anything,” a chat interface for non-developers to use the same infrastructure directly. Feedback welcome — especially from people building agents, copilots, or AI workflow systems. https://ift.tt/xGTdvkC
New ask Hacker News story: AI Is Boring
AI Is Boring
3 by sverp | 1 comments on Hacker News.
Everyone is building the same thing. Does anyone know of something genuinely exciting happening in AI?
3 by sverp | 1 comments on Hacker News.
Everyone is building the same thing. Does anyone know of something genuinely exciting happening in AI?
New ask Hacker News story: Should I go open-source or give free trial and then charge the users?
Should I go open-source or give free trial and then charge the users?
2 by akarshhegde18 | 2 comments on Hacker News.
Built a product that runs on your machine, understands your work and takes care of your Jira task updates for you. No data leaves your machine
2 by akarshhegde18 | 2 comments on Hacker News.
Built a product that runs on your machine, understands your work and takes care of your Jira task updates for you. No data leaves your machine
New ask Hacker News story: What Can't Fable Do?
What Can't Fable Do?
3 by alimhaq | 0 comments on Hacker News.
I'm curious, has anyone encountered scenarios where Fable wasn't able to solve a problem, fix a bug, complete some task, etc. outside of the (questionable) restrictions in place? Every conversation I read about Fable devolves into either two conclusions: (1) Fable is insane and was able to do X in one or two prompts or fix issues that other models and humans had been grinding on for a while or (2) Fable sucks because my prompt triggered the guardrails. But I haven't seen much of "Fable wasn't able to do X" and in my work so far I haven't seen it really fail or mess up much at all.
3 by alimhaq | 0 comments on Hacker News.
I'm curious, has anyone encountered scenarios where Fable wasn't able to solve a problem, fix a bug, complete some task, etc. outside of the (questionable) restrictions in place? Every conversation I read about Fable devolves into either two conclusions: (1) Fable is insane and was able to do X in one or two prompts or fix issues that other models and humans had been grinding on for a while or (2) Fable sucks because my prompt triggered the guardrails. But I haven't seen much of "Fable wasn't able to do X" and in my work so far I haven't seen it really fail or mess up much at all.
New Show Hacker News story: Show HN: CLI that helps AI agents avoid vulnerable dependencies
Show HN: CLI that helps AI agents avoid vulnerable dependencies
2 by modelorona | 0 comments on Hacker News.
deptrust is a CLI that checks package versions for known vulnerabilities across npm, PyPI, crates.io, Go modules, RubyGems, NuGet, Maven, Packagist, pub.dev, CocoaPods, Hex.pm, Hackage, GitHub Actions, and more. It runs locally as a CLI and as an MCP server. It calls public package registry and OSV APIs directly; there is no hosted deptrust service. I built this because AI coding agents kept suggesting outdated or vulnerable package versions. I kept having to manually tell tools like Claude and Codex to use newer, safer versions. deptrust gives the agent a quick way to verify whether a dependency version has known vulnerabilities before it installs or recommends it. You can install it with: 1. pnpx @clidey/deptrust@latest install 2. brew install clidey/tap/deptrust 3. Or directly with go: go install github.com/clidey/deptrust/cmd/deptrust@latest
2 by modelorona | 0 comments on Hacker News.
deptrust is a CLI that checks package versions for known vulnerabilities across npm, PyPI, crates.io, Go modules, RubyGems, NuGet, Maven, Packagist, pub.dev, CocoaPods, Hex.pm, Hackage, GitHub Actions, and more. It runs locally as a CLI and as an MCP server. It calls public package registry and OSV APIs directly; there is no hosted deptrust service. I built this because AI coding agents kept suggesting outdated or vulnerable package versions. I kept having to manually tell tools like Claude and Codex to use newer, safer versions. deptrust gives the agent a quick way to verify whether a dependency version has known vulnerabilities before it installs or recommends it. You can install it with: 1. pnpx @clidey/deptrust@latest install 2. brew install clidey/tap/deptrust 3. Or directly with go: go install github.com/clidey/deptrust/cmd/deptrust@latest