Hacker News
Streisand Effect AI used by Police exists for maximum provocation and entrapment
They hated Jesus because he told them the truth. Also see: The Streisand Effect of Flagged Hacker News Posts: https://www.youtube.com/watch?v=rkhrf5MNYRk
I will not have my posts be flagged. I will be heard.
Comments URL: https://news.ycombinator.com/item?id=43891265
Points: 1
# Comments: 0
How your mouth could be killing your heart
Article URL: https://theconversation.com/how-your-mouth-could-be-killing-your-heart-254860
Comments URL: https://news.ycombinator.com/item?id=43891248
Points: 1
# Comments: 0
Show HN: My AI Native Resume
I've been deeply involved in working with AI agents and large language models (LLMs) for a while now. During a recent job search, I found myself repeatedly explaining my skills and experiences to various assistants. Around the same time, I was creating content for my website to help hiring teams understand my capabilities better and make informed decisions.
MCP had started to gain momentum and I saw a way to reduce my toil. So I built an MCP server that can effectively communicate my qualifications as a job candidate. This server acts as an AI-powered resume, providing an understanding of my professional background and a set of tools, prompts and resources to help explore my skills and experiences.
The code is open source, so you can create your own AI-driven resume server. Check it out here: https://github.com/jhgaylor/node-candidate-mcp-server.
During my job search I paired my mcp server with others such as notion, hirebase, and gmail to build a leads database, write cover letters, and track my job search.
Comments URL: https://news.ycombinator.com/item?id=43891245
Points: 2
# Comments: 0
Show HN: LLM-Exe – A Modular TypeScript Toolkit for LLM Application Development
I've recently updated llm-exe, a modular TypeScript library built specifically to simplify creating applications with Large Language Models (LLMs). The package allows you to call LLM's from various providers without changing the underlying code.
The library focuses on a structured, lightweight, modular design that lets developers easily assemble complex workflows from reusable components:
Prompts: Structured templating for managing sophisticated prompts with built-in Handlebars support.
Parsers: Components that transform raw LLM responses (strings) into structured data, supporting JSON, arrays, enum extraction, and custom parsing logic.
LLM Providers: Abstracted interfaces for various providers, including OpenAI, Anthropic, xAI, Google Gemini, AWS Bedrock, and Ollama, allowing seamless switching without changing implementation logic.
Executors: The LLM executor takes an llm, a prompt, optionally a parser, and wraps in a well-typed function. An LLM executor is a container that can be used to call an LLM with a pre-defined input and output; with additional values provided at the time of execution. An LLM executor's input and output types are determined by the prompt and parser respectively.
Utilities: Helpers for common tasks such as prompt debugging, caching, and managing conversation state.
Here's how you might create a structured executor:
--- import { createLlmExecutor, createChatPrompt, createParser, useLlm } from 'llm-exe';
const llm = useLlm("openai.gpt-4o-mini");
const prompt = createChatPrompt<{input: string}>('Translate the phrase "{{input}}" into French, Spanish, and German, returning the results as an unordered markdown list.');
const parser = createParser('listToArray');
const translateExecutor = createLlmExecutor({ llm, prompt, parser });
// result is typed as string[]
// the input is also well-typed!
const result = await translateExecutor.execute({ input: "Hello, world!" });
console.log(result);
// Outputs: ["Bonjour le monde!", "¡Hola, mundo!", "Hallo, Welt!"] ---
Check out the documentation and more examples here: llm-exe.com. I'd love your feedback or contributions!
Comments URL: https://news.ycombinator.com/item?id=43891162
Points: 1
# Comments: 0
Wikidive – AI guided rabbitholes in Wikipedia
Article URL: https://wikidive.tulv.in/
Comments URL: https://news.ycombinator.com/item?id=43891147
Points: 3
# Comments: 2
Self-driving cars can talk to each other while on the road
Unparalleled Misalignments
Article URL: https://rickiheicklen.com/unparalleled-misalignments.html
Comments URL: https://news.ycombinator.com/item?id=43891128
Points: 1
# Comments: 0
Show HN: AgentShield SDK – Runtime security for agentic AI applications
Hi HN,
We built AgentShield, a Python SDK and CLI to add a security checkpoint for AI agents before they perform potentially risky actions like external API calls or executing generated code.
Problem: Agents calling arbitrary URLs or running unchecked code can lead to data leaks, SSRF, system damage, etc.
Solution: AgentShield intercepts these actions:
- guarded_get(url=...): Checks URL against policies (block internal IPs, HTTP, etc.) before making the request.
- safe_execute(code_snippet=...): Checks code for risky patterns (os import, eval, file access, etc.) before execution.
It works via a simple API call to evaluate the action against configurable security policies. It includes default policies for common risks.
Get Started:
Install: pip install agentshield-sdk
Get API Key (CLI): agentshield keys create
Use in Python: from agentshield_sdk import AgentShield # shield = AgentShield(api_key=...) # await shield.guarded_get(url=...) # await shield.safe_execute(code_snippet=...)
Full details, documentation, and the complete README are at https://pypi.org/project/agentshield-sdk/
We built this because securing agent interactions felt crucial as they become more capable. It's still early days, and we'd love to get your feedback on the approach, usability, and policies.
Comments URL: https://news.ycombinator.com/item?id=43891111
Points: 2
# Comments: 1
P1 Modula-2 for Mac
Article URL: https://modula2.awiedemann.de/
Comments URL: https://news.ycombinator.com/item?id=43891100
Points: 2
# Comments: 0
A structured coalescent model reveals deep ancestral structure shared by humans
Article URL: https://www.nature.com/articles/s41588-025-02117-1
Comments URL: https://news.ycombinator.com/item?id=43891069
Points: 1
# Comments: 0
Goodput
Article URL: https://en.wikipedia.org/wiki/Goodput
Comments URL: https://news.ycombinator.com/item?id=43891044
Points: 1
# Comments: 0
A Fifth of American Adults Can't Read
Article URL: https://www.thefp.com/p/a-fifth-of-american-adults-cant-read-i-teach-them
Comments URL: https://news.ycombinator.com/item?id=43891043
Points: 6
# Comments: 1
I built a tool that lets you cross-post your social media anywhere
Article URL: https://reposter.social
Comments URL: https://news.ycombinator.com/item?id=43891031
Points: 1
# Comments: 2
My Blog Lasted Longer Than the Confederacy (2024)
Article URL: https://soatok.blog/2024/07/21/my-furry-blog-has-lasted-longer-than-the-confederacy/
Comments URL: https://news.ycombinator.com/item?id=43890998
Points: 3
# Comments: 1
Before Our Attention Was a Commodity: Memories of a Pre-Web Internet
Article URL: https://blog.sanfranciscan.org/2025/05/04/before-our-attention-was-a-commodity/
Comments URL: https://news.ycombinator.com/item?id=43890973
Points: 1
# Comments: 0
Balcony solar has proven popular in Germany, but U.S. isn't ready
Article URL: https://www.bdcnetwork.com/home/news/55287607/balcony-solar-has-proven-wildly-popular-in-germany-but-us-isnt-ready
Comments URL: https://news.ycombinator.com/item?id=43890908
Points: 1
# Comments: 0
Ansible Now() Function
Article URL: https://www.ansiblepilot.com/articles/mastering-time-in-ansible-the-now-function
Comments URL: https://news.ycombinator.com/item?id=43890906
Points: 1
# Comments: 0
LLM ported to the C64
Article URL: https://hackaday.com/2025/05/03/llm-ported-to-the-c64-kinda/
Comments URL: https://news.ycombinator.com/item?id=43890901
Points: 1
# Comments: 0
Flash USDT; Instant Transfer Utility
Join the Revolution in Crypto Transactions Today! Don’t miss out on the opportunity to elevate your crypto trading experience. With Flash USDT, you can enjoy stealth-level speed and complete security — making it a must-have for anyone serious about cryptocurrency trading.
How It Works in Just Three Steps:
Choose the Receiver Wallet: Enter the wallet address you want the USDT to appear in. Set Your Flash Parameters: Decide how much USDT to send and how long you want it to stay visible. Flash Instantly: Initiate the flash and watch your USDT appear in seconds. After the timer ends, it vanishes — leaving no traces.
Discover the future of crypto transactions with Flash USDT today!
Pricing Plans;
We offer various pricing plans to fit your trading needs. Choose the one that suits you best:
$200 for $2000 Flash USDT Tradable Flash Balance Instant Use Upon Delivery $400 for $5000 Flash USDT Tradable + Transfer-Ready Bonus: Wallet Cloaking Mode $1000 for $15,000 Flash USDT Elite Flash Balance Cold Wallet Stealth Injection Lifetime Access to Updates Flash Access Packages: Starter License: $700 Flash 1 Wallet Daily 500 USDT Daily License Last for 5 Days Pro License: $1200 (Most Popular) Flashes 3 Wallets Daily 2000 USDT Daily Priority Telegram Support Elite Plan: $2500 Unlimited Daily Flashes Up to 25,000 Daily Lifetime Telegram Access
Unlock your crypto potential and act now — packages are limited!
Connect with Us! Have any questions? We’re here to help! Chat with a verified seller on Telegram for instant assistance.
@drewztooolz +1 (770) 666–2531
Comments URL: https://news.ycombinator.com/item?id=43890897
Points: 1
# Comments: 0
List of remote jobs updated daily
Article URL: https://listofremotejobs.com/
Comments URL: https://news.ycombinator.com/item?id=43890856
Points: 1
# Comments: 2