Tristan

[Tristan @exectristan]

@dylayed Are there common use cases where attaching multiple sandboxes (or multiple agents, even) to a single drive is valuable?

[Tristan @exectristan]

@zechengzh Yes, exactly that!

[Tristan @exectristan]

@zechengzh How much of a concern is race conditions?

[Tristan @exectristan]

@leodev @e2b @daytonaio @upstash @boatdotdev @boxd_sh @freestyle_dev What benchmarks are you planning to run?

[Vishnu Krishnaprasad @an_engineer_log]

there's more sandbox companies in sf than homeless 😅 😂

[Tristan @exectristan]

Got the free pizza. Then had Hark email support to delete my data.

[Hark @hark_labs]

We're giving away $1,000,000 of pizza right now all across the U.S.

It's pizza night, on us. To celebrate the launch of Hark Pro. Download the app and tell Hark to order a free pizza. https://t.co/YRHAFYfxKd

Instructions in thread to order yours now.

[Tristan @exectristan]

@hark_labs Is this just a ploy to harvest all our data?

[Tristan @exectristan]

Terribly needed, we don’t have enough sandbox providers as it is.

[Perplexity @perplexity_ai]

We're open sourcing SPACE, the sandbox platform that runs every task in Perplexity Computer.

SPACE gives each agent task its own Firecracker microVM and controls credentials and outbound traffic from outside the VM.

Available now under Apache 2.0: https://t.co/DzSmq6o1xA

[Tristan @exectristan]

I swing between ‘Wait, this is basically AGI’ and ‘How did it mess up something this simple?’ about 20 times a day.

[Tristan @exectristan]

@leodev @DevinAI @hoplite_sh @capydotai Devin is unusable for me. 6.1 Luna burns through my $200/mo Devin subscription faster than 6 Astra burns through my $200/mo Codex Pro Max subscription.

[Tristan @exectristan]

Another day, another new agent sandbox provider.

[InsForge @insforge]

InstaBox is live ☁️

We’re opening up access in batches, invite-only, as we scale up compute.

Request an invite code on Hang’s post below, and we'll be in touch 👇

[Tristan @exectristan]

@HeyGarrison @HeyGarrison Curious how you think about cache locality with ComputeSDK Actions.

Blacksmith gets a lot of its CI speed from keeping Docker/build caches close to its own runner fleet. If the same workflow can move across Blaxel, etc., how do you avoid losing that locality?