Jev powered "auto mode" in the Temporal Agent Harness is such a natural fit. Jev is a big unlock for efficient agent oversight. Very excited to see how straightforward it was to integrate along existing harness seams. More to come here for sure.
This is the wrong approach
Running agents inside the sandbox is simple in theory
But in production, the sandbox's blast radius becomes your agent's blast radius
For example, with the DigitalOcean architecture:
- Sandbox crashes / OOMs → agent loop dies with it
- Disk fills up or env breaks → agent is bricked
- Exposing your own API as tools → needs extra infra
- Creds can't live in the sandbox → needs a whole credential proxy layer
- Sleeping sandbox can't wake itself for crons
- Updating agent code → updating the binary in every VM
- Monitoring → the sandbox has to report its own failures
Separate the harness from the sandbox instead:
- Run the harness in your backend
- Expose the sandbox as tools
- Keep creds, history, and permissions outside
- Wake the sandbox only when needed
Amp, Vercel Eve, Cloudflare Flue, Claude Managed Agents, and OpenAI's Agents SDK all use this architecture for a reason
I wrote more on why and how here: rivet.dev/blog/2026-07-27-ru…
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: do.co/4ysh3it
Standalone Activities are now GA.
Use Temporal as a job queue for background work. Each job is recorded before it reaches a Worker and retried according to policy. Its status, errors, and result stay in one place.
All the details are on our blog: temporal.io/blog/a-better-jo…
Dario Amodei is right that safeguards need time to catch up with AI capabilities. But companies deploying agents face a related question: how much authority should they hand over?
When model output becomes real action, AI safety doesn’t stop at the model.
temporal.io/blog/ai-safety-d…
We are incredibly excited to announce Temporal's Series E.
It wasn't so long ago that we met Max and Samar working on open source Cadence at Uber.
Fast forward and they're at 4K+ paying customers, a $12.55B valuation, 200%+ NDR 🔥
I still remember when Lenny first flew to Seattle to convince Samar and Max that they had built something with such enormous potential. His conviction in their project and their capacity for leadership may have exceeded their own. And yet Temporal continues to defy even our expectations!
We are doubling down on @temporalio as we lead their $550M Series E raise.
Agents now carry out real work across the systems businesses depend on: booking travel, shipping code, moving money. A single job can cross services and clouds and run for hours or days, and every boundary is another place for the work to fail.
We believe @temporalio has become the de facto solution to this problem.
We’re thrilled to be backing co-founders @SamarAtTemporal and @mfateev and the entire Temporal team again.
CC: @anoushkavaswani, @ravi_lsvp, @buckymoore, @jshen99
We are doubling down and leading @temporalio’s $550M Series E Raise.
Agents now do real work: calling a payments API, updating a record, closing out a job. Temporal is the foundational infrastructure helping enable these actions. We believe Temporal is table stakes for any company leveraging AI.
Over 4,300 paying customers. More than $250M in run-rate revenue. It has been a pleasure to work with @SamarAtTemporal and @mfateev and the entire @temporalio team. Excited to see you take the company to new heights.
Cc: @ravi_lsvp, @buckymoore, @jshen99
A special day for Temporal. We raised $550M at a $12.55B valuation.
Thank you to the team building Temporal and the developers who keep pushing us to make it better. temporal.io/blog/temporal-ra…
Temporal @temporalio is hiring for 64 roles across engineering, product, sales, marketing, customer support and more.
They provide infrastructure for critical AI systems at OpenAI, Cursor, Replit and Lovable. OpenAI handles millions of requests with Temporal, while Cursor’s cloud coding agents run 50M+ actions a day across 7M+ workflows.
Founders @SamarAtTemporal and @mfateev spent years building workflow systems at Amazon, Microsoft and Uber. Temporal is remote-first, with openings across six countries.
We've sponsored Crystal Palace's men's team since earlier this year. Starting this weekend, we're extending that to @cpfc_w; same club, same commitment.
More on why: temporal.io/blog/temporal-pa…
If you'll be in SF a week from today, join us. Coffee, smoothies and pastries are on us. Come just for a quick bite / to say hi, or stay to chat and cowork!
luma.com/cafe-cowork-tempora…
@ShrivuShankar great to see Temporal adding value. Engineers building state management, retries, DLQ, etc as agents become more distributed and takes action in real world is not going to scale. It's so exciting to see @temporalio becoming the reliable foundation for AI systems.
At our annual conference, Replay, we had two nights and one workday to flash, assemble, and charge 2,300 conference badges before they hit the stage. We used Temporal to run the coordination: 32 badges per carrier, one Workflow per badge, and retries that ran on their own.
Read the full story: temporal.io/blog/ignition-fl…
When a Pub/Sub event hits your subscriber, what happens next usually turns into retry logic, status tables, and a reconciliation job you now have to monitor.
How a Standalone Activity replaces all of that: temporal.io/blog/bridging-pu…