From 83cf7a770702617ee00b972e98b873dac44c6b3d Mon Sep 17 00:00:00 2001 From: shanshan wang Date: Sun, 3 May 2026 16:44:46 -0500 Subject: [PATCH] add tensorlake in sandbox provider list --- content/docs/core-concepts.mdx | 2 +- content/docs/index.mdx | 2 +- content/docs/run-jobs/cloud-sandboxes.mdx | 4 ++-- content/docs/run-jobs/results-and-artifacts.mdx | 3 ++- 4 files changed, 6 insertions(+), 5 deletions(-) diff --git a/content/docs/core-concepts.mdx b/content/docs/core-concepts.mdx index e3469ec..f9299a4 100644 --- a/content/docs/core-concepts.mdx +++ b/content/docs/core-concepts.mdx @@ -19,7 +19,7 @@ An [agent](/docs/agents) is a program that completes tasks. Agents are defined b ## Container environment -Environments in Harbor are containers, typically defined as Docker images using a `Dockerfile`. The `BaseEnvironment` interface provides a unified interface for interacting with environments. Many cloud container runtimes are already supported out of the box, including [Daytona](https://www.daytona.io/), [Modal](https://modal.com/), [E2B](https://e2b.dev/) and [Runloop](https://runloop.ai/). Other container runtimes can be supported by implementing the `BaseEnvironment` interface. +Environments in Harbor are containers, typically defined as Docker images using a `Dockerfile`. The `BaseEnvironment` interface provides a unified interface for interacting with environments. Many cloud container runtimes are already supported out of the box, including [Daytona](https://www.daytona.io/), [Modal](https://modal.com/), [E2B](https://e2b.dev/), [Runloop](https://runloop.ai/) and [Tensorlake](https://www.tensorlake.ai/). Other container runtimes can be supported by implementing the `BaseEnvironment` interface. ## Trial diff --git a/content/docs/index.mdx b/content/docs/index.mdx index c31ccd1..a84a560 100644 --- a/content/docs/index.mdx +++ b/content/docs/index.mdx @@ -14,5 +14,5 @@ Harbor provides: - Simple, modular interfaces for environments, agents, and tasks - All popular CLI agents pre-integrated - A registry of popular benchmarks and datasets -- Integrations with cloud sandbox providers like [Daytona](https://www.daytona.io/), [Modal](https://modal.com/), [E2B](https://e2b.dev/) and [Runloop](https://runloop.ai/) for horizontal scaling +- Integrations with cloud sandbox providers like [Daytona](https://www.daytona.io/), [Modal](https://modal.com/), [E2B](https://e2b.dev/), [Runloop](https://runloop.ai/) and [Tensorlake](https://www.tensorlake.ai/) for horizontal scaling - Integrations with frameworks like SkyRL and GEPA for optimizing agents diff --git a/content/docs/run-jobs/cloud-sandboxes.mdx b/content/docs/run-jobs/cloud-sandboxes.mdx index d0093db..a2f8f51 100644 --- a/content/docs/run-jobs/cloud-sandboxes.mdx +++ b/content/docs/run-jobs/cloud-sandboxes.mdx @@ -9,7 +9,7 @@ Using a cloud sandbox provider shifts command execution to the cloud, making tri ## Using a cloud sandbox provider -There are many cloud sandbox providers to choose from. We recommend [Daytona](https://www.daytona.io/), because we have found them to be the most flexible. Other good options are [Modal](https://modal.com/), [E2B](https://e2b.dev/) and [Runloop](https://runloop.ai/). +There are many cloud sandbox providers to choose from. We recommend [Daytona](https://www.daytona.io/), because we have found them to be the most flexible. Other good options are [Modal](https://modal.com/), [E2B](https://e2b.dev/), [Runloop](https://runloop.ai/) and [Tensorlake](https://www.tensorlake.ai/). ```bash harbor run -d "" \ @@ -25,4 +25,4 @@ We run up to 100 trials in parallel on a MacBook Pro with 14 cores. Daytona supports multi-container deployments. To use multi-container tasks with Daytona, include an `environment/docker-compose.yaml` file in your task definition. -Other cloud sandbox providers (Modal, E2B, Runloop) do not currently support multi-container environments. For those providers, you will need to use single-container tasks or switch to Daytona or the local Docker environment. +Other cloud sandbox providers (Modal, E2B, Runloop and Tensorlake) do not currently support multi-container environments. For those providers, you will need to use single-container tasks or switch to Daytona or the local Docker environment. diff --git a/content/docs/run-jobs/results-and-artifacts.mdx b/content/docs/run-jobs/results-and-artifacts.mdx index 099c420..5932ad8 100644 --- a/content/docs/run-jobs/results-and-artifacts.mdx +++ b/content/docs/run-jobs/results-and-artifacts.mdx @@ -7,7 +7,7 @@ Harbor can automatically collect files from the sandbox environment after each t ## Convention directory (zero configuration) -Any files written to `/logs/artifacts/` inside the sandbox are collected automatically with no configuration needed. For Docker environments, this directory is volume-mounted directly to the host. For remote environments (Daytona, Modal, E2B, etc.), files are downloaded after the trial finishes. +Any files written to `/logs/artifacts/` inside the sandbox are collected automatically with no configuration needed. For Docker environments, this directory is volume-mounted directly to the host. For remote environments (Daytona, Modal, E2B, Tensorlake, etc.), files are downloaded after the trial finishes. For example, if your task's test script or agent writes files to `/logs/artifacts/`: @@ -130,3 +130,4 @@ Artifact collection works across all environment types: | Daytona | Downloaded after trial | Downloaded after trial | | Modal | Downloaded after trial | Downloaded after trial | | E2B | Downloaded after trial | Downloaded after trial | +| Tensorlake | Downloaded after trial | Downloaded after trial |