frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Show HN: Agensi – Curated marketplace for AI agent skills (SKILL.md)

https://www.agensi.io
1•Samuellrose•1h ago
Hi HN. I'm Samuel, founder of Agensi (agensi.io). I'm a non-technical founder in the Netherlands. Built the platform mostly with Claude Code and Lovable over the last few months.

What it is: a curated marketplace for SKILL.md skills. SKILL.md is the folder-plus-instructions format Anthropic shipped for teaching AI coding agents (Claude Code, Cursor, Codex CLI, Gemini CLI, and others) new capabilities. Creators publish skills, users install them into their agent. Every listing goes through an automated security scan before going live.

Two things I've tried to do that felt missing from the ecosystem:

1. Curation with security. Running someone else's code on your machine without review is a real risk, especially given the ToxicSkills and ClawHavoc research that came out earlier this year (36% of sampled skills had prompt injection vectors). Every skill on Agensi runs through an automated scan checking permission boundaries, outbound network requests, dependency red flags, and common malware patterns. It's not perfect and doesn't replace reading the code yourself, but it filters the obvious bad stuff.

2. Creator monetization. Two paths, creator picks whichever fits the skill. Direct sales: set a price, buyer pays once, creator keeps 80% minus a $0.50 per-sale fee. MCP subscription pool: 70% of net MCP subscription revenue distributed monthly based on which skills paying users actually pulled in via MCP. The idea is that time spent encoding your workflow knowledge as a SKILL.md shouldn't just earn you GitHub stars.

There's also an MCP server that makes skill discovery agent native. Four tools: search_skills, get_skill, submit_skill_request, get_skill_requests. Your agent can search the catalog and pull in a skill mid-task instead of you browsing the site. If the skill is not on the marketplace, the agent can add a skill request that pings every creator on the platform. Subscription is $9/month with a 3-day trial.

Honest state of things:

- 40 creators, 200+ skills listed - 66 articles in the Learn section, ~7k active users over the past 28 days - MCP subscription launched last week. 1 paying subscriber so far (yearly), 1 trialist who cancelled before converting. Too early to say anything about conversion. - Pre-seed discussions with VC’s ongoing - GMV month 1 of ~200$

Things I'd genuinely want feedback on:

- Creator split. Direct 80/20 and MCP 70/30. I picked numbers that felt defensible after running the math on infrastructure and payment costs. Open to the argument that either should skew further toward creators.

- Security scan depth. The current scan is automated and covers the basics. Whether to add creator identity verification, manual review queues, or public scan reports per skill is an open question for me.

- Moat. Anthropic, Cursor, and others could launch first-party skill stores whenever they want. My thesis is that a curated cross-agent marketplace is defensible because first-party stores are always vendor-locked. I would like to hear counterarguments.

- Skills you've wished existed but couldn't find. Open to hearing what's missing from the catalog.

Browse is free, no signup needed: https://agensi.io

I'll be in the comments.

What 23K vulnerabilities reveal about audit report quality in Web3

https://colab.research.google.com/drive/1Wp4yyEmXYjHATak7Bmy2lf6DNgyjlxgI?usp=sharing
1•zaevlad•3m ago•1 comments

Show HN: Vibe coding t-rex.world

1•nirupan23•5m ago•0 comments

SQLite WAL in 2026: an architecture note on checkpoint starvation

https://cronfeed.work/oss-sqlite-wal-architecture-note-checkpoint-starvation-concurrency-boundary...
1•tonyl•6m ago•0 comments

Social Credit

https://nicholasdecker.substack.com/p/social-credit
1•barry-cotter•7m ago•0 comments

Elon vs. Altman: What Their Infrastructure Stacks Reveal About Power

https://mythcoreops.substack.com/p/elon-vs-altman-what-their-infrastructure
1•ReadTheProducts•9m ago•0 comments

Tim Cook's Impeccable Timing

https://stratechery.com/2026/tim-cooks-impeccable-timing/
3•hasheddan•10m ago•0 comments

AI Database Landscape in 2026: Vector, ML-in-DB, LLM-Augmented, Predictive

https://aito.ai/blog/the-ai-database-landscape-in-2026-where-does-structured-prediction-fit/
1•arauhala•14m ago•1 comments

Show HN: Dreamtime – A fresh bedtime story for your children every night

https://trydreamtime.com
2•jblox•15m ago•0 comments

How well do LLMs work outside English? We tested 8 models in 8 languages [pdf]

https://info.rws.com/hubfs/2026/trainai/llm-data-gen-study-2.0-campaign/trainai-multilingual-llm-...
1•curioussquirrel•15m ago•2 comments

Polymarket in talks to raise money at about $15B valuation

https://www.reuters.com/business/polymarket-talks-raise-money-about-15-billion-valuation-informat...
2•JumpCrisscross•16m ago•0 comments

Show HN: Hevc.js – HEVC/H.265 to H.264 transcoding in WebAssembly (236KB)

https://www.hevcjs.dev/
1•privaloops•16m ago•0 comments

Datahugging shields proprietary AI models from research that could disprove them

https://www.nature.com/articles/s44387-026-00094-2
2•apwheele•19m ago•0 comments

My file access workaround for cron in Tahoe

1•noduerme•21m ago•0 comments

Using LLMs effectively isn't about prompting

https://www.seangoedecke.com/beyond-prompting/
1•adityaathalye•23m ago•0 comments

Smoking ban for people born after 2008 in the UK agreed

https://www.bbc.co.uk/news/articles/cn08jy6w0l5o
7•AndrewDucker•24m ago•3 comments

Claude Sonnet 4.6 thinking duplicates what it has said, wasting tokens

1•szmarczak•25m ago•0 comments

French prosecutors issue summons to Musk after raid on X offices in Paris

https://courthousenews.com/french-prosecutors-issue-summons-to-musk-after-raid-on-x-offices-in-pa...
7•vrganj•25m ago•1 comments

AI Prevents Meritocracy

https://www.privacyguides.org/videos/2026/04/19/interview-with-carissa-veliz-author-of-privacy-is...
2•freddyym•25m ago•0 comments

Global growth in solar "the largest ever observed for any source"

https://arstechnica.com/science/2026/04/global-growth-in-solar-the-largest-ever-observed-for-any-...
2•AndrewDucker•26m ago•0 comments

Java script error: why The Devil Wears Prada 2's Starbucks tie-in leaves a taste

https://www.theguardian.com/film/2026/apr/21/the-devil-wears-prada-2-starbucks-collaboration
2•TruffleLabs•27m ago•2 comments

California tries to fix its housing mess

https://economist.com/united-states/2025/10/14/california-tries-to-fix-its-housing-mess
1•andsoitis•28m ago•0 comments

Digging for clues about the North Pole's past

https://www.technologyreview.com/2026/04/21/1135231/digging-for-truth-north-pole-past-seabed/
1•joozio•31m ago•0 comments

Schmoozebots: Study finds flattery will get AI everywhere

https://www.theregister.com/2026/04/20/chatbots_win_trust_by_sounding/
1•pseudolus•31m ago•1 comments

Ask HN: Is taking input from your microphone a TRNG?

1•smalltorch•32m ago•0 comments

BlackRock's AI Transformation

https://www.wsj.com/cio-journal/inside-blackrocks-ai-transformation-03a1e8c7
1•monkeydust•32m ago•0 comments

Quantum 'Jamming' Explores the Fundamental Principles of Nature

https://www.quantamagazine.org/quantum-jamming-explores-the-truly-fundamental-principles-of-natur...
1•pseudolus•32m ago•0 comments

Laws of Software Engineering

https://lawsofsoftwareengineering.com
7•milanm081•35m ago•2 comments

Show HN: Rapunzel, a tree-style tab browser for agents

https://github.com/salmanjavaid/rapunzel
1•WasimBhai•35m ago•0 comments

My favorite hard science fiction books by old masters

https://bookdna.com/best-books/hard-science-fiction-by-old-masters
2•bwb•36m ago•1 comments

AI Affiliate Campaign Builder

https://www.indiehackers.com/product/ai-affiliate-campaign-builder
1•rooseveltc•37m ago•0 comments