frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Show HN: Integrations gateway for agents with 2FA for destructive ops (OSS)

https://github.com/yakkomajuri/agentport
2•yakkomajuri•1h ago
Hey HN!

I've been wanting to use something like OpenClaw for a while but couldn't get myself to give it access to anything important due to all the risks involved. Prompt injection is still a problem (even though some people seem to ignore it) and so are hallucinations and mishaps that cause agents to do things like delete production data [1].

Even harnesses like Claude Code and Codex are subject to this, particularly since we're getting progressively looser about how we run them e.g. Conductor is really popular and runs agents without any sandboxing.

That means we're in a bit of an all-or-nothing situation. There are people who just ignore the risks and connect everything to their agents and reap benefits from it while being subject to more risk, and there are others that just don't connect anything because they are mindful of the potential issues.

I've been quite cautious but have wanted to run more autonomous agents and so I built the component I needed to enable me to do so: AgentPort.

AgentPort is a gateway that connects to any service (e.g. Gmail, GitHub, Stripe, PostHog, Linear) and let's you set granular permissions for what the agent can do automatically, what it needs your approval for, and what it can never do.

For example, you can set `list_customers` and `get_customer` on the Stripe integration to "Auto-approve" but `create_refund` to "Ask for approval". The agent will thus be able to do a lot in the background independently but when it comes to a potentially destructive operation it will be blocked and receive an approval link to send to you. You can then approve or deny the call with those exact parameters e.g. `create_refund(customer_id: 1234, amount: 12)`.

Agents connect via MCP or CLI and have access to all the integrations you connected without ever getting API keys. Kind of like Composio but with granular permissions and open source.

The goal with AgentPort is to specifically address two vulnerabilities that agents are subject to:

1. Destructive operations on downstream services: It can't delete a database unless you explicitly approve 2. Credential exfiltration: Your agent never sees API keys

AgentPort also helps with sensitive data exfiltration, but that is more nuanced and complicated to defend against if the agent has an internet connection [2].

Ultimately, AgentPort was the missing piece for me to start running more autonomous agents that have access to third-party services, and hopefully it can unlock use cases for you too. There's a ton more work needed around securing agents (Claws in particular) and I've both been writing about it [3] and intend to do more in this space, so if you're thinking about similar things let's have a chat.

The repo is https://github.com/yakkomajuri/agentport and you can run it locally with docker compose in a minute or use the one-liner install to deploy a prod instance (domain, TLS, etc.) in just a few mins as well.

[1] "An AI agent deleted our production database. The agent's confession is below" (https://news.ycombinator.com/item?id=47911524) [2] See my post "On agents dropping production databases": https://yakko.dev/blog/on-agents-dropping-production-dbs [3] https://yakko.dev/blog

Authsome – open-source local auth proxy for AI agents

https://github.com/manojbajaj95/authsome
1•pkhodiyar•31s ago•1 comments

NIST's NVD Changes: A Wake-Up Call for CVE-Driven Security

https://www.oligo.security/blog/nists-nvd-changes-a-wake-up-call-for-cve-driven-security
1•curmudgeon22•1m ago•0 comments

Auto Polo

https://en.wikipedia.org/wiki/Auto_polo
1•canjobear•1m ago•0 comments

From Convergence to Confidence: Push-Button Verification for RDTs

https://kcsrk.info/verification/rdts/lean/2026/04/28/from-convergence-to-confidence/
1•ibobev•1m ago•0 comments

Superwhisper now integrates with Claude Code

https://superwhisper.com/claude-code
2•Neeeks•2m ago•1 comments

Physicists Discover the Most Complex Forms of Ice Yet

https://www.quantamagazine.org/physicists-discover-the-most-complex-forms-of-ice-yet-20260427/
1•ibobev•3m ago•0 comments

Circular Arc Approximation

https://www.johndcook.com/blog/2026/04/28/circular-arc-approximation/
1•ibobev•3m ago•0 comments

Google and Pentagon reportedly agree on deal for 'any lawful' use of AI

https://www.theverge.com/ai-artificial-intelligence/919494/google-pentagon-classified-ai-deal
1•granzymes•3m ago•0 comments

Arietta: A framework for creating local AI voice assistants w. knowledge, tools

https://github.com/robert-mcdermott/arietta-voice
1•mcdermott•4m ago•0 comments

Lapsus$ – Data Repository

https://lapsus.by/
1•speckx•5m ago•0 comments

Introduction to Quantum Information Science Lecture Notes (2018) [pdf]

https://www.scottaaronson.com/qclec.pdf
1•Tomte•7m ago•0 comments

IMPulse: Open-source incident manager (Grafana OnCall alternative)

https://impulse.bot/
1•ditsi•7m ago•1 comments

Elephants, Goldfish and the New Golden Age of Software Engineering

https://drensin.medium.com/elephants-goldfish-and-the-new-golden-age-of-software-engineering-c336...
1•fridek•8m ago•0 comments

I Built an AI Trading Platform in Six Days. That's Terrifying

https://www.bloomberg.com/opinion/articles/2026-04-28/ai-trading-bots-are-creating-a-major-financ...
1•wslh•9m ago•1 comments

You're probably taking the wrong painkiller

https://asteriskmag.substack.com/p/youre-probably-taking-the-wrong-painkiller
1•speckx•9m ago•0 comments

One common charging solution for all

https://single-market-economy.ec.europa.eu/sectors/electrical-and-electronic-engineering-industri...
2•doener•10m ago•0 comments

The Neuroscience Behind Writing: Handwriting vs. Typing [pdf]

https://pmc.ncbi.nlm.nih.gov/articles/PMC11943480/
1•thunderbong•11m ago•0 comments

Meta Is Preparing to Have to Undo Its Manus Acquisition After China Ban

https://www.wsj.com/tech/ai/meta-is-preparing-to-have-to-undo-its-manus-acquisition-after-china-b...
2•Brajeshwar•11m ago•0 comments

The Best (Query) Plans of Mice and Men

https://ohadravid.github.io/posts/2026-04-query-plans/
1•ohr•12m ago•0 comments

Initial Commit: A Dedication

https://www.khanna.law/blog/initial-commit
3•hkhanna•13m ago•0 comments

Claude for Creative Work

https://www.anthropic.com/news/claude-for-creative-work
3•l1n•14m ago•0 comments

Show HN: CV Mirror – see what 5 ATS parsers do to your CV

https://cv-mirror-web.vercel.app/
1•GS_Projects•14m ago•0 comments

Software engineering may no longer be a lifetime career

https://www.seangoedecke.com/software-engineering-may-no-longer-be-a-lifetime-career/
1•twapi•16m ago•0 comments

Meta, Microsoft look to trim workforces amid heavy AI spending

https://fortune.com/2026/04/23/meta-microsoft-layoffs-job-cuts-not-filling-open-roles-voluntary-b...
2•gmargari•18m ago•0 comments

Show HN: SuperVoiceMode dictation experiment became an AI voice interface

https://voicemode.io/
1•mikezx•20m ago•0 comments

Iran's Meme War Against Trump Ushers in a Future of 'Slopaganda'

https://www.nytimes.com/interactive/2026/04/25/business/iran-trump-israel-war-memes.html
3•ryan_j_naughton•21m ago•0 comments

Scroll-Driven Animations: Exploring the majestic new animation-timeline API

https://www.joshwcomeau.com/animation/scroll-driven-animations/
1•joshwcomeau•22m ago•0 comments

Train collision in Indonesia kills 14 as rescuers work to reach survivors

https://www.theguardian.com/world/2026/apr/28/indonesian-train-crash-fatal
1•Imustaskforhelp•22m ago•0 comments

KiCad Breadboard Builder

https://github.com/kerstensrobin/kicad-breadboard
1•radeeyate•23m ago•0 comments

Kdts: A TypeScript compiler with type-driven optimizations

https://github.com/KimlikDAO/kdts
1•szaglam•23m ago•0 comments