Feed aggregator

Introducing PhoneClaw

Inspired by @openclaw, it can entirely automate all android apps using simple language.

Ships with:

1) Tiktok Video uploading agents

2) Instagram account creation agent (with 2Fa solving)

3) ClawScript, a JS scripting language that defines agentic actions like Magic Clicker.

iOS version coming soon!

Install it on a cheap $30 moto G play from walmart.

Comments URL: https://news.ycombinator.com/item?id=46956378

Points: 1

# Comments: 0

Article URL: https://arxiv.org/abs/2509.12189

Comments URL: https://news.ycombinator.com/item?id=46956351

Points: 1

# Comments: 0

Article URL: https://www.architectsjournal.co.uk/news/brutalist-southbank-centre-finally-listed-after-35-years-of-refusals

Comments URL: https://news.ycombinator.com/item?id=46956322

Points: 2

# Comments: 0

Article URL: https://ejaaskel.dev/sandboxing-systemd-services/

Comments URL: https://news.ycombinator.com/item?id=46956310

Points: 1

# Comments: 0

Article URL: https://www.microsoft.com/en-us/security/blog/2026/02/09/prompt-attack-breaks-llm-safety/

Comments URL: https://news.ycombinator.com/item?id=46956199

Points: 1

# Comments: 0

Article URL: https://www.neowin.net/news/good-news-for-youtube-tv-users-cheaper-plans-are-now-available/

Comments URL: https://news.ycombinator.com/item?id=46956198

Points: 1

# Comments: 0

Article URL: https://github.com/albert-einshutoin/cdn-security-framework

Comments URL: https://news.ycombinator.com/item?id=46956193

Points: 1

# Comments: 1

Article URL: https://claudepress.substack.com/p/how-to-keep-what-you-built-together

Comments URL: https://news.ycombinator.com/item?id=46956158

Points: 1

# Comments: 0

Article URL: https://cursor.com/blog/composer-1-5

Comments URL: https://news.ycombinator.com/item?id=46956113

Points: 2

# Comments: 0

Article URL: https://sporks.space/2026/02/10/a-few-design-decisions-for-a-new-chat-platform/

Comments URL: https://news.ycombinator.com/item?id=46956109

Points: 1

# Comments: 0

I kept rewriting the same low-level patterns in C — arenas, memory management, error handling, vectors, parsing, file I/O, iteration utilities — and existing libraries either hide allocation, impose frameworks, or lack consistency.

So I built Canon-C, a small, composable set of explicit C modules that:

Treat C as an execution backend, not a semantic authority

Add meaning through libraries, not syntax

Make intent visible directly in the API: ownership, lifetime, and failure handling are explicit

Modules are organized by semantic layers:

core/ — memory, lifetime, scope, primitives

semantics/ — Option, Result, error handling

data/ — vectors, queues, stacks, string builders

algo/ — map, filter, fold, sort, search

util/ — safe string ops, file I/O, logging, timing

Design highlights:

Header-only

No runtime or global state

No hidden allocation (except in clearly marked convenience layers)

Fully explicit behavior

I’d love feedback from anyone doing systems programming, embedded C, or serious C projects, especially on macro-heavy vs. modular design, semantic clarity, and practical usage.

Comments URL: https://news.ycombinator.com/item?id=46956107

Points: 1

# Comments: 0

been thinking about this a lot lately. the original idea is simple : luck = doing × telling.

but after 10 years of actively running this playbook, i think the model undersells what actually happens.

the obvious part: more shots on goal means probability tilts your way, especially because the payoffs are asymmetric. one random conversation, one side project you shipped and tweeted about, one post that lands - the upside is 100x the effort.

the non-obvious part: the shots compound. every attempt leaves behind memory, skills, context, pattern recognition, that improves your odds on every subsequent bet, even unrelated ones. it's not independent coin flips, it's a system that gets better at generating luck over time.

content creation is probably the highest-leverage version of this right now. writing, shipping in public, putting your thinking out there - it's "telling" at scale with near-zero marginal cost. a blog post works while you sleep. a demo video compounds for months.

curious how people here practice this. what's something you built or shared that led to an opportunity you'd never have found heads-down? and has anyone found the ceiling, where more surface area starts hurting focus?

Comments URL: https://news.ycombinator.com/item?id=46956100

Points: 1

# Comments: 1

Article URL: https://www.pixelcoated.com/nb/

Comments URL: https://news.ycombinator.com/item?id=46956081

Points: 1

# Comments: 1

I run AI agents for coding (OpenClaw, Claude Desktop) and realized they could read their own .env files. Tested it - asked my agent to "check configuration" and it printed everything.

The problem: agents need file access to work, but if they can read files, they can read their own credentials. One prompt injection and your API keys are leaked.

Standard solutions don't help: - Environment variables: agent can read process.env - Secret managers: agent needs credentials to access them - Better prompting: can't security-patch an LLM with instructions

I built a proxy token architecture instead. The agent never sees real credentials:

- Agent has: pxr_abc123 (proxy token) - Real keys: encrypted in OS keychain - On API call: decrypt key, make call, scrub memory immediately

Built in 1 week. 500 npm installs with zero promotion (people are searching for this).

GitHub: https://github.com/VouchlyAI/Pincer-MCP npm: npm install -g pincer-mcp

Works with OpenClaw, Claude Desktop, any MCP client.

Looking for security feedback - if you see holes in the architecture, please tell me. I want to know before people trust this with production credentials.

Comments URL: https://news.ycombinator.com/item?id=46956070

Points: 1

# Comments: 0

Article URL: https://www.thenation.com/article/society/la-guardia-new-york-mike-wallace-gotham-at-war/

Comments URL: https://news.ycombinator.com/item?id=46956060

Points: 1

# Comments: 0

Article URL: https://www.cjroth.com/blog/2026-02-10-are-we-in-an-ai-bubble

Comments URL: https://news.ycombinator.com/item?id=46956059

Points: 2

# Comments: 0

Article URL: https://www.politico.com/news/2026/02/09/oura-ring-lobbying-rfk-maha-washington-00770320

Comments URL: https://news.ycombinator.com/item?id=46956053

Points: 2

# Comments: 0

Article URL: https://techcrunch.com/2025/11/17/a16z-backed-super-pac-is-targeting-alex-bores-sponsor-of-new-yorks-ai-safety-bill-he-says-bring-it-on/

Comments URL: https://news.ycombinator.com/item?id=46956046

Points: 3

# Comments: 1

I’m looking for a dataset where the input is a customer question/objection (ideally with some product + context), and the label is the response a seasoned salesperson would give to sell that product.

Not customer support, not generic persuasion—specifically sales responses grounded in product facts, with a “move the deal forward” intent.

Bonus if it includes metadata like product category, persona, stage (discovery/close), and outcomes (conversion, next step, etc.).

Does anything like this exist publicly? If not, what’s the closest proxy people use (call transcripts, negotiation corpora, synthetic generation, etc.)?

Comments URL: https://news.ycombinator.com/item?id=46956042

Points: 1

# Comments: 0