Adil Hafeez

[Chris Kim @chriskim_dev]

I just automated deal room acquisition screening with @Box & @digitalocean Managed Agents.

Drop a new file in the deal room → Triggers the sandbox agent to wake up and rebuild the review memo with the latest info.

No more stale deal room memos.

Full setup deep dive video coming soon.

[Salman Paracha @salman_paracha]

This was such a fun proof of concept to see in person! Go @Box (@levie ) and @digitalocean (@paddix )

[DigitalOcean @digitalocean]

Action Gateway is the half of Managed Agents that handles tools. 16,000+ of them, from 400+ providers, behind one endpoint.

Ask to "check if the server is still alive." The tool is called ping_network_endpoint. With no words in common, keyword search makes that tool hard to find.

@r_o_connor explains how Action Gateway finds the right one anyway, 99.3% of the time.

Try it in public preview: https://t.co/o1kdhkzVNX

[DigitalOcean @digitalocean]

DigitalOcean Managed Agents is now in public preview.

Run Claude Code, Codex, or your own LangGraph agent in a runtime environment that pauses when idle. Put its tools behind one governed endpoint, and pick from 75+ open and proprietary models. One cloud, one bill.

Prompts to get started available in the blog: https://t.co/QQqQ9w28lG

[Avi Chawla @_avichawla]

https://t.co/sgTW6eRQTw

[Salman Paracha @salman_paracha]

A year in the making. Thanks to @twilio and @Atlassian for their feedback as we built this with them. This is our take on the elusive problem of routing between LLMs: A preference-aligned approach that offers a practical mechanism to capture subjective evaluation criteria. https://t.co/OOiSY3OhOc

[AK @_akhaliq]

Meet Arch: The AI-native proxy server for agents. Why do you need a proxy? So you can move faster by focusing on core business objectives of your AI apps.

Arch handles the pesky heavy lifting in building agentic systems. -- offers fast ⚡️ query understanding and routing, seamless integration of prompts with APIs to build agentic tasks, and unified access and observability of LLMs.

[Salman Paracha @salman_paracha]

3 weeks ago we had the #1 trending function calling LLM on @huggingface. Today we have the #1 AND #2 trending function calling LLM. Fast, cheap, and SOTA performance packaged in https://t.co/jptSy14rJK to build fast agentic apps. Grateful to the community rallying behind our work. cc @_akhaliq @ClementDelangue

[Salman Paracha @salman_paracha]

Super thrilled that in such a short time - especially for an infrastructure project - we achieved the 1000+ @github stars ⭐️ milestone. Love the community building around https://t.co/9gprrTVxLF. Onwards and upwards 🚀

[Rust Trending @RustTrending]

katanemo / archgw: Arch is an intelligent gateway for agents. Engineered with (fast) LLMs for the secure handling, rich observability, and seamless integration of prompts with your APIs - all outside business logic. Built by the core ... ★928 https://t.co/Ldl3ufMqGt

[Salman Paracha @salman_paracha]

I am thrilled that we are now the #1 🏆 trending function calling LLM on @huggingface - the fastest, most efficient models with performance to match frontier LLMs. Arch-Function is engineered in Arch https://t.co/jptSy14rJK - the intelligent gateway for agents - we'd❤️ for you to try...

[Salman Paracha @salman_paracha]

I promised @samuelcolvin that we'd integrate with logfire in our 0.2.0 release. I lied, we pulled that in for 0.1.6. With zero lines of application code, developers who want rich out-of-the-box tracing for agents (prompt, tools call, LLM) - can use Arch + @pydantic Logfire.

[AK @_akhaliq]

Arch

Build fast, observable, and personalized AI agents.

Arch is an intelligent gateway for agents. Engineered with (fast) LLMs for the secure handling, rich observability, and seamless integration of prompts with your APIs - all outside business logic. Built by the core contributors of Envoy proxy, on Envoy.

[Salman Paracha @salman_paracha]

This is a post about how we thought differently about agents and function-calling. Below is @_philschmid version of how function-calling works. Its' accurate, as flow control goes from the app to an LLM, followed by a series of exchanges. But...

[Salman Paracha @salman_paracha]

I created a short video about how Arch (the intelligent proxy for agents built on Envoy) can help with intent routing for smarter retrieval, guardrails for safety, and function calling for agent personalization via simple @FastAPIs. Of course, @Gradio is our natively integrated chat interface, and you can select from @MistralAI and @OpenAI LLMs for summarization (more coming soon)

[Peter Jausovec @pjausovec]

I'll be doing a live stream tomorrow and explore the Arch gateway.

Join me at 3pm PST. Link below.

[Adil Hafeez @adilhafeez]

@pjausovec Hey Peter - @salman_paracha and I have updated archgw to 0.1.5 with some updates. Please sync the repo. It has some improvements (related to docker) and some bug fixes related to prompt guards and enums.

https://t.co/mnL72zk0zM

[Tom Dörr @tom_doerr]

"Arch is an intelligent prompt gateway. Engineered with (fast) LLMs for the secure handling, robust observability, and seamless integration of prompts with your APIs - outside business logic. Built by the core contributors of Envoy proxy, on Envoy."

[Salman Paracha @salman_paracha]

Build production-ready agentic apps using @Gradio + @FastAPIs + Arch : the intelligent infrastructure primitive for gen AI https://t.co/Hp1WXEqAZx

Arch is a proxy for agents - Engineered with (fast) LLMs for the secure handling, robust observability, and seamless integration of prompts with APIs - all outside business logic.

[Salman Paracha @salman_paracha]

https://t.co/ao4D3Y0UX7 - would love your support ❤️

Please consider upvoting and commenting on our launch