Feed aggregator
After Microsoft's AI overreach, Gentoo begins its march away from GitHub
Experiential Reinforcement Learning
Article URL: https://arxiv.org/abs/2602.13949
Comments URL: https://news.ycombinator.com/item?id=47061995
Points: 2
# Comments: 0
If you're always listening to an audiobook, you're not alone
Article URL: https://www.washingtonpost.com/books/2026/02/11/audiobook-listening-trends/
Comments URL: https://news.ycombinator.com/item?id=47061978
Points: 1
# Comments: 1
Show HN: Mock any HTTP request from DevTools, with AI-generation and zero setup
Hi HN,
I built this after using Requestly, Mokku, Mockiato, Tweak, and Mockoon. Each one either paywalled the features I actually needed, required a separate server running on my machine, or just didn't fit the way I work.
The browser is already open. DevTools is already open. I wanted the mocking to live there too, not in a separate app I have to remember to start.
So roughly a month ago, I started building my own tools. It intercepts network requests at the browser level and lets you swap the response, status code, headers, body without touching your code or spinning up anything. Right-click a in the Network tab, mock it, done. v1.0.3 is out now and it's been stable enough that I've been using it daily on client work for months.
A few things I care about that shaped the decisions:
- Local-first by default. Your mocks live in the browser's storage.
- The free tier is actually free. Unlimited mocks, unlimited projects, OpenAPI import, GitHub Sync, BYOK AI (OpenAI-compatible or local Ollama). No trial countdown, no credit card.
Currently it is only support Chromium-based browser. The things I'm still considering are supporting other browser also, such as Firefox, Opera, etc.
Extension: https://chromewebstore.google.com/detail/nilcfehohhhfjekckib...
If you've built something in this space or had opinions on why existing tools fell short, I genuinely curious.
Comments URL: https://news.ycombinator.com/item?id=47061971
Points: 1
# Comments: 0
Show HN: Poncho, a general agent harness built for the web
Hi HN! I'm building poncho, a framework for building custom AI agents that are version-controlled in git, developed locally, and deployed as isolated endpoints (serverless-friendly by default).
Poncho agents follow the same conventions as openclaw or claude code so they probably feel familiar.
You can talk to them via terminal or web ui to build new skills or configure the agent, and they're compatible with the Agent Skills open spec, so you can port your skills. Small caveat: right now it's compatible with .md skills and js/ts scripts, but many skills are bash or python-based. I'm still figuring out what would be the best way to extend support for those skills without overcomplicating it, if anybody has any ideas I appreciate it!
Some features:
- Git-native: agent behavior, skills, and tests live in your repository (reviewable diffs + easy rollbacks). - Single-file agent definition: define runtime config + instructions in AGENT.md (YAML frontmatter + prompt content). - Skills you can ship: AgentSkills-style skills/*/SKILL.md plus TypeScript/JavaScript scripts under scripts/. - MCP support: connect remote tool servers and inject required environment variables through config. - Conversation-first API + streaming: stored conversations with SSE streaming responses and tool events. - Pluggable storage + memory: local files for dev or hosted stores (e.g. Upstash), with optional persistent memory + recall. - Testing + observability: poncho test workflows and OpenTelemetry traces/events.
Now the cool thing about poncho agents is that they're super easy to deploy to Vercel/Fly/Lambda/etc, so you can share them with anybody in your team and enable non-technical people with custom skills.
I also built and deployed a couple example agents here: - https://github.com/cesr/product-agent - https://github.com/cesr/marketing-agent
I'd love some feedback, I started building poncho because I wanted a fast and easy way to build and share agents with my team. Let me know what you think!
Comments URL: https://news.ycombinator.com/item?id=47061949
Points: 2
# Comments: 0
The Sun's Temper Tantrums: What You Should Know About Solar Storms
Google's Pixel 10A Looks Stylish for a Low-Cost Flagship Phone
My Pixel 10A Hands-On: Google Adds Fast Charging, Satellite SOS for $499
India Tells University To Leave AI Summit After Presenting Chinese Robot as Its Own
Cogent Security Raises $42 Million for AI-Driven Vulnerability Management
The Series A funding round, led by Bain Capital, brings the total raised by Cogent to $53 million.
The post Cogent Security Raises $42 Million for AI-Driven Vulnerability Management appeared first on SecurityWeek.
Identity-aware SSH proxy using your OS native SSH client
Article URL: https://www.pomerium.com/docs/capabilities/native-ssh-access
Comments URL: https://news.ycombinator.com/item?id=47061423
Points: 1
# Comments: 0
You Only Think They Work for You
Article URL: https://steveblank.com/2026/02/18/you-only-think-they-work-for-you/
Comments URL: https://news.ycombinator.com/item?id=47061415
Points: 2
# Comments: 0
Trump moves closer to a major war with Iran
Article URL: https://www.axios.com/2026/02/18/iran-war-trump-military-strikes-nuclear-talks
Comments URL: https://news.ycombinator.com/item?id=47061407
Points: 1
# Comments: 0
DPO, your simplest RL pipeline with two rollouts
Article URL: https://fireworks.ai/blog/dpo-as-rl
Comments URL: https://news.ycombinator.com/item?id=47061405
Points: 1
# Comments: 0
The End of Local
Article URL: https://charlielabs.ai/blog/the-end-of-local/
Comments URL: https://news.ycombinator.com/item?id=47061401
Points: 2
# Comments: 0
Show HN: Browser Terminal Use – run local CLI/agent loops in browser terminals
I built Browser Terminal Use: a Chrome extension + local daemon + CLI that lets you run commands in a browser-hosted terminal from your local shell on macOS,
Why I built it: setups). - I wanted a bridge that keeps local automation while executing remotely in the browser terminal context.
How it works:
- browterm CLI sends exec requests to browterm-daemon on localhost. - Daemon serializes requests (single active command) and routes to a bound Chrome terminal tab. - Command is wrapped with markers to extract clean output + remote exit code. - Supports timeout and cancel.
Current limitations:
- Browser terminal websocket protocols vary by vendor. - Some UIs can block input fallbacks. - Cross-origin iframe terminals can reduce observability.
I’d really value feedback on reliability across terminal vendors and ideas for improving compatibility.
Comments URL: https://news.ycombinator.com/item?id=47061398
Points: 1
# Comments: 1
The flaws in India's AI plans
Article URL: https://www.economist.com/asia/2026/02/17/the-flaws-in-indias-ai-plans
Comments URL: https://news.ycombinator.com/item?id=47061374
Points: 1
# Comments: 1
There Is No Great Millennial Novel
Article URL: https://substack.magazinenongrata.com/p/why-the-great-millennial-novel-doesnt
Comments URL: https://news.ycombinator.com/item?id=47061357
Points: 1
# Comments: 1
Show HN: Clawlet – AI agent with built-in semantic memory, one binary
Clawlet is a personal AI agent that ships as a single, self-contained binary. No runtime, no package manager, no external database. The main thing that sets it apart: built-in hybrid semantic memory search (vector similarity + full-text) using a bundled SQLite with vector extensions. The index is just a local .sqlite file — no separate vector DB to run. Drop the binary on any machine and memory search just works.
GitHub: https://github.com/mosaxiv/clawlet
Comments URL: https://news.ycombinator.com/item?id=47061356
Points: 1
# Comments: 0
