Code & AI agent sandboxes: read the guide

microsandbox

Open coreSandbox

A computer for anything, anywhere. Open-source microVM sandboxes for AI agents and untrusted code.

Homepage Docs Pricing Repository

Scanned 12 Oct 2026 · Schema v5 · features v21

From: $49/mo
Monthly plan plus metered usage; private-beta rates
Free tier: Not documented
Cloud Free: $0 monthly hard-capped pool; invitation access required. Separate from free Apache-2.0 self-host runtime.
Isolation: Virtual machine
hypervisor, or nothing at all
Regions: —
Not disclosed on the public surface.
Capabilities: 17 / 61
6 partial · 32 fail · 0 not assessed
Agent readiness: 11/11 · 8/11
read · drive, pre-login surface

Advertised for

AI agents, untrusted user code/plugins, CI jobs, developer environments, scraping and automation in own kernel microVMs.

Built for

Developers, then technical

· CLI/SDK-first local microVM quickstart; Rust/TypeScript/Python/Go/Ruby APIs.
· Untrusted code, OCI images, own kernels and explicit local/cloud backends.

01 Pricing Monthly plan plus metered usage; private-beta rates

Tier	EntryFree	Builder	Teams	TopEnterprise
Price	$0/month	$49/month	$299/month	Custom committed/BYOC
Storage	1 GB	50 GB	250 GB	500 GB
vCPU hours	5 h	500 h	2000 h	4000 h
Memory gb hours	20 h	2000 h	8000 h	16000 h
Storage overage	—	$0.15/GB	$0.15/GB	—
vCPU hours overage	—	$0.05/h	$0.05/h	—
Memory gb hours overage	—	$0.0162/h	$0.0162/h	—

Free tier: Cloud Free: $0 monthly hard-capped pool; invitation access required. Separate from free Apache-2.0 self-host runtime.

vCPU-hours, memory GiB-hours, writable disk GiB-hours, durable storage GiB-hours; compute billed in seconds.·Card required: unconfirmed·microsandbox.dev

02 Runtime Each fact carries its source; dots show confidence. Open a row for the evidence.

Isolation & lifecycle

Isolation primitive

Virtual machinehypervisor, or nothing at all

Persistence

Persistent sandboxes and managed named/default directory volumes; cloud paths resolve against organization host storage, not the client laptop.

Snapshots / forks

Yes

Resources & limits

Storage

Managed persistent directory volumes and writable root disks; cloud files accessible through common SDK path operations.

Limits & quotas

Private beta: Free 3 concurrent/2 persistent, Builder 1,000 concurrent/50 persistent, Teams 25,000 concurrent/500 persistent. Free sandbox shape 1 vCPU/512 MiB/8 GiB disk; includes 5 vCPU-hours/20 memory GiB-hours/20 disk GiB-hours and 1 GiB durable-storage cap.

GPU

Access

Deploy interface

container-imageclisdkapimcp

SDK languages

RustTypeScriptPythonGoRuby

Network & secrets

Egress controls

Yes

Secrets

Host-side, destination-bound placeholder substitution; allowed endpoints receive the actual credential.

Bring your own cloud

Evaluation: same open runtime in your cloud or on your metal; you operate it and vendor helps evaluate/deploy/stay current.

Placement & use

Jurisdiction

Unknown — hosted placement/data residency reviewed with the team before rollout.

Advertised workloads

AI agents, untrusted user code/plugins, CI jobs, developer environments, scraping and automation in own-kernel microVMs.

03 Capabilities 17 pass · 6 partial · 32 fail · 61 of 61 assessed

Supported — Documented on the public surface and verified.
Partially supported — Exists with caveats: gated, thin, or enterprise-only.
Not supported — Checked and not offered.
Not assessed — Scanned before this capability was tracked. Not a lack.

Compute & storage 3/11

Persistent diskStorage that survives restart/stop (not ephemeral-only).

Supported

Object storageManaged S3-compatible / blob object store the platform PROVIDES to a user's workloads (buckets, presigned URLs). DISTINCT from persistent-disk (block/volume attached to an instance) and from a self-host dependency (source.self_host.deps[]=["object-storage"] is an infra dep to RUN an OSS tool — the opposite direction). SCOPE varies and matters: native first-class (Railway Buckets, Cloudflare R2) vs a sibling SKU / extension (Fly Tigris, Bunny Storage, DigitalOcean Spaces) — record which in the evidence quote / note. Admitted v15.

Not supported

Snapshots / forkingSnapshot or copy-on-write fork of a running instance (filesystem-only counts). Whether the snapshot also restores memory + running processes is recorded separately by memory-snapshot (v19).

Supported

Live-state (memory) snapshotsA snapshot / checkpoint / pause captures a running instance's MEMORY and PROCESS state and restores it, so running processes continue where they left off instead of a freshly booted instance with the same files. Counts whether user-triggered (snapshot, checkpoint, pause-to-storage) or used to restore pre-initialised state (memory snapshots for fast cold starts). Distinct from snapshots-forks (any snapshot or fork, including filesystem-only; a product passing that key may still fail this one) and from scale-to-zero (an idle-cost property; suspension may or may not keep memory). Actor multiplexing (see scale-to-zero) records its live-state half here — only when the suspended actor's MEMORY/process state is what comes back, not a freshly booted worker with the actor's files.

Partially supported

Scale to zeroIdle instances suspend at no / negligible cost. ACTOR MULTIPLEXING (many suspended stateful actors sharing a small pool of pre-warmed workers, woken on request — Agent Substrate, AX, kagent; Rivet's hibernating actors) has NO key of its own (proposal REJECTED v20) — score the idle-cost half HERE and the live-state half under memory-snapshot; the oversubscription ratio / density is a performance characteristic the registry does not key.

Partially supported

GPU computeGPU-backed instances available.

Not supported

Multi-regionCompute placeable in more than one geographic region.

Not supported

Sandbox templates / custom imagesBuild reusable sandbox/VM images from an OCI image or Dockerfile (a base you fork instances from).

Supported

Cloud-init / ignition configCustomise a VM / microVM / instance at boot via cloud-init (or ignition) user-data — the same boot-metadata interface as AWS/Azure/GCP. Inject packages, users, SSH keys, files, and run-commands declaratively at launch. Distinct from sandbox-templates (a PREBUILT image you fork instances from) — this is PER-LAUNCH declarative customisation, not a baked image.

Not supported

Desktop / computer-use sandboxA virtual desktop (GUI) for LLM computer-use — browser + desktop automation, often via noVNC/VNC.

Partially supported

Selectable isolation runtimeThe user chooses the isolation runtime per workload / environment / sandbox from more than one option (e.g. microVM vs gVisor vs plain container, native namespaces vs gVisor, gVisor vs VM). Records the CHOICE; the isolation_primitive facet still records the best-evidenced primary. A platform that switches runtime automatically without a user choice (e.g. gVisor forced for GPU) does not pass on that alone.

Not supported

Networking & delivery 1/8

Automatic HTTPSTLS endpoints provisioned automatically per workload.

Not supported

Built-in reverse proxyPlatform routes external traffic to workload services.

Not supported

Custom domainsBring-your-own domain via CNAME/DNS.

Not supported

Public preview URLsShareable per-deploy / per-branch URLs.

Not supported

Direct public IPv6Per-instance stable public IPv6 for raw TCP/UDP (beyond a managed HTTPS proxy) — e.g. mosh / raw sockets.

Not supported

Egress controlsConfigurable OUTBOUND network policy — egress firewall / allowlist / egress proxy on traffic LEAVING a workload. Scoped to OUTBOUND only (v15): private-network REACHABILITY between a user's own services is private-networking, not this. Static-egress-IP belongs here (an outbound concept).

Supported

Private networkingA platform-internal private network between a user's OWN services — service-to-service connectivity, VPC, service mesh, and/or service discovery, not exposed to the public internet. DISTINCT from egress-controls (OUTBOUND firewall/allowlist), reverse-proxy (PUBLIC inbound ingress), and direct-ipv6 (per-instance PUBLIC addressing). Admitted v15.

Not supported

Email send / receiveSend and/or receive email from the workload.

Not supported

Edge KV storeA globally-replicated key-value store read from WITHIN an edge runtime at the PoP (low-latency, eventually-consistent) — Cloudflare Workers KV, Deno KV, Fastly KV Store, Akamai EdgeKV, Vercel Edge Config. Distinct from first-party-db (a managed DATABASE service you provision and connect to) and from http-query-api (HTTP access to an external store) — edge-kv is colocated with the edge compute and read at the edge.

Not applicable

Developer access 6/9

Public APIDocumented programmatic API.

Supported

OpenAPI specMachine-readable OpenAPI / Swagger spec published.

Supported

CLIOfficial command-line tool.

Supported

SDKOfficial client library/libraries.

Supported

SSH accessDirect SSH into workloads.

Supported

Filesystem APIRemote file operations over the API (list/read/write/watch/copy/upload) without SSH.

Supported

Git deployDeploy from a Git repo / push-to-deploy.

Not supported

Private container registryHosted private image registry.

Not supported

E2B-compatible APIDrop-in compatibility with the E2B sandbox API/SDKs — existing E2B client code runs against this product by changing the endpoint/domain + key, with no business-code changes. E2B itself passes as the reference implementation. Distinct from public-api / sdk (the product has its own API — this key says it also speaks another vendor's) and from self-host-oss (a self-hosted E2B-compatible server is scored on both).

Not supported

HTTP data APIQuery / read / write the stored DATA over plain HTTP/REST (or GraphQL) WITHOUT the native protocol driver (Postgres wire / Redis RESP) — agent- and edge-friendly data access via fetch(). The DATA plane; distinct from public-api (a control-plane / management API). Auto-generated-from-schema REST/GraphQL (PostgREST, pg_graphql) is a flavour of this.

Not applicable

AI & agents 3/11

Built-in coding agentShips a first-party AI coding agent in the product.

Not supported

Agent skillsInstallable / declarable agent skills or tools.

Supported

AI app builderPrompt-to-app / vibe-coding generation surface.

Not supported

MCP serverExposes a Model Context Protocol server for agents.

Supported

MCP clientActs as an MCP client / consumes MCP tools.

Not supported

MCP-server hostingRun/host MCP servers as a workload on the platform (distinct from exposing a first-party MCP control server — that is mcp-server).

Not supported

LLM gatewayRoutes LLM API calls through the platform (multi-provider).

Not supported

BYOK for LLMsBring-your-own LLM provider keys.

Not supported

Model inference API (first-party)A FIRST-PARTY hosted LLM/model inference API — the provider serves its OWN catalogue of (usually open-weight) models behind a (typically OpenAI-compatible) endpoint, billed per token. Distinct from llm-gateway (a multi-PROVIDER proxy/router for YOUR calls to external providers like OpenAI/Anthropic) and from byok-llm (bring your own provider keys) — here the platform IS the inference provider, serving its own models.

Not applicable

Autonomous QA agentA first-party AUTONOMOUS AI agent that QA-TESTS a running app / implementation — exercising it like a real user (clicking through flows, checking behaviour) and reporting bugs/defects (often back to a coding agent, or before a human review). Distinct from built-in-agent (a code-AUTHORING agent — this one VERIFIES the result, it does not write the feature; a product can ship both), from ai-guided-code-review (organises a diff for a HUMAN to review — not an autonomous tester), and from observability (runtime metrics/logs, not active testing).

Not supported

Credential-injection proxyThe platform holds real secrets OUTSIDE the workload and injects them on the wire (an egress / broker proxy swaps a placeholder or phantom token for the real credential on matching outbound requests), so code running in the sandbox — e.g. an untrusted agent — never sees or can exfiltrate the real key. Distinct from secrets-management (storage + injection INTO the workload as env vars/files — the workload holds the secret) and from egress-controls (outbound allow/deny POLICY; a product can have an allowlist without injecting credentials, and vice versa).

Supported

Preinstalled coding agentThe platform installs and launches a THIRD-PARTY coding agent (a harnesses.yml entry — Claude Code, Codex, OpenCode, Pi ...) inside a workspace/sandbox it provisions, with the agent's auth wired in (your own account or API key), so you start an agent session rather than an empty box. Distinct from built-in-agent (a FIRST-PARTY agent the vendor ships) and from harnesses[] (which records that a harness CAN target the product, not that the product runs it for you).

Partially supported

Data 0/1

First-party databasePlatform-provided managed database.

Not supported

Database branchingCopy-on-write branch of a DATABASE (schema + data) as a dev-lifecycle primitive — branch your data like code, usually instant via a metadata operation. Distinct from snapshots-forks (forks a running compute INSTANCE / VM state) and from preview-environments (a full per-PR environment that may USE a DB branch). For a data product a CoW branch maps HERE, not to snapshots-forks.

Not applicable

Vector searchFirst-class storage, indexing, and similarity search of vector embeddings — a native vector type/index or a managed vector DB. The data-layer primitive for RAG / semantic retrieval.

Not applicable

Ephemeral instant DB (no-signup)Provision a working, usable instance via a single command / HTTP call with NO account, signup, or card — typically time-boxed (expires unless claimed). A zero-onboarding path an agent can hit directly.

Not applicable

Platform & operations 0/6

Scheduled tasksCron / scheduled job primitive.

Not supported

Durable workflowsDurable execution / long-running workflow primitive.

Not supported

Preview environmentsA full EPHEMERAL environment spun up per PR / branch — compute + addons + (often) a DB branch — torn down on merge. More than preview-urls (which is just a shareable URL for one deploy).

Not supported

Built-in observabilityFirst-party metrics / logs / tracing / APM / profiling included in the platform (no third-party stack required to see what a workload is doing).

Partially supported

Managed CI/CD runnersPlatform-hosted, drop-in CI/CD runner VMs (GitHub Actions / GitLab), typically faster and cheaper than the CI vendor's own hosted runners, billed per runner-minute; isolated + ephemeral per job. Distinct from git-deploy (push-to-deploy of an APP) and scheduled-tasks (cron) — this is managed CI COMPUTE you point your existing pipelines at.

Not supported

Project template marketplaceA marketplace / gallery of prebuilt PROJECT STARTER templates (apps or websites) a user clones or launches a new project from and then customises — the "start from a template" surface of a builder / PaaS. Distinct from sandbox-templates (a COMPUTE base IMAGE you fork runtime INSTANCES from — Dockerfile / OCI), from app-builder (the prompt-to-app GENERATION surface — a product can have both), and from one-click OSS-app catalogues / helm-addon-marketplace (deploy a FINISHED app as-is, not a starter you build on).

Not supported

Organisation & security 4/9

Built-in authPlatform-provided authentication / IAM for workloads.

Not supported

SSO / SAMLEnterprise single sign-on.

Supported

Secrets managementFirst-class secrets storage / injection.

Supported

TeamsMulti-user team accounts.

Supported

SharingShare a running workload/resource via link (optionally gated).

Not supported

BYOCRuns inside the customer's own cloud account.

Partially supported

Deploy into customers' VPCsA SaaS vendor uses the platform to deploy and operate ITS software inside ITS CUSTOMERS' cloud accounts / VPCs (nested or "customer BYOC"), managed from the vendor's control plane. Distinct from byoc (your own workloads in your own cloud account).

Not supported

Workload identityThe platform gives each workload its own verifiable identity (OIDC token, SPIFFE SVID, cloud IAM federation) that it exchanges for short-lived credentials, so no long-lived static secret is stored or injected. Distinct from secrets-management (stores / injects static secrets) and credential-injection-proxy (swaps real secrets in on the wire).

Not supported

Self-host / OSS coreOpen-source core that can be self-hosted.

Supported

04 Agent readiness Pre-login surface · counts, not a grade

Can an agent read it?

parse-ability · docs, metadata, machine-readable surfaces

11/11

11 full · 0 partial · 100% of applicable checks

llms.txt
Supported
“Apex and docs llms.txt fetched; docs index has ~300 doc links.”
llms-full.txt
Supported
“Apex full text 51,051 bytes; comprehensive docs full text 2,260,469 bytes retained as pointer.”
robots.txt AI policy
Supported
“Apex robots explicitly allows GPTBot, ClaudeBot, anthropic-ai, PerplexityBot and others.”
JSON-LD
Supported
“Homepage Organization and SoftwareApplication JSON-LD.”
Sitemap
Supported
“Apex sitemap.xml fetched with nontrivial page list.”
Meta tags
Supported
“Descriptive title, description and Open Graph/Twitter metadata in fetched homepage HTML.”
Markdown negotiation
Supported
“Mintlify .md content variants fetched for quickstart, cloud, API overview and SDKs.”
Content efficiency
Supported
“HTML 112527 bytes; extracted readable text 5856 bytes (5.20%); script/style omitted.”
Docs discoverable
Supported
“First-class homepage docs/quickstart links; GitHub homepageUrl and README point back to docs.”
Machine-readable pricing
Supported
“Server-rendered private-beta tiers/rates with numeric pools; uncertainty about promotion noted separately.”
No anti-bot wall
Supported
“Homepage, pricing, docs markdown and linked OpenAPI fetched without challenge. Two specific source failures are recorded separately.”

Can an agent drive it?

act-ability · APIs, SDKs, MCP, credentials, webhooks

8/11

8 full · 0 partial · 73% of applicable checks

Public API
Supported
“Cloud REST API and sandbox control SDKs documented.”
OpenAPI spec
Supported
“Two valid docs-hosted OpenAPI 3.1.0 specs fetched; pointers retain size/provenance.”
MCP server
Supported
“Documented stdio MCP package plus apex MCP server card.”
SDK packages
Supported
“Five official SDKs: Rust/TS/Python/Go/Ruby.”
Auth documented
Supported
“API overview documents org msb_ API keys, personal msb_pat_ tokens and Authorization: Bearer.”
Self-serve credentials
Not supported
“Private beta requires human invitation/manual approval. API-key and personal-token creation exists for authenticated users; not unrestricted agent signup.”
Webhooks
Not supported
“No signed webhooks found; audit events are polled REST resources.”
CLI tool
Supported
“Official msb CLI with cloud backend selector.”
Error model
Supported
“API overview shows JSON error.code, error.message and optional error.details.”
Rate limits documented
Not supported
“Organization/resource quotas disclosed; no API rate-limit response-header/backoff contract found.”
Agent config files
Supported
“Dedicated AI-agent docs, installable skills, stdio MCP and well-known server card/skills index.”

05 Coding harnesses Which agents and IDEs are documented to run here

Advertised integrations 9

Claude Code Anthropic
Supported
“claude mcp add --transport stdio microsandbox -- npx -y microsandbox-mcp; Agent Skills 'Works with Claude Code, Cursor, Codex, Gemini CLI, GitHub Copilot'.”
Source docs.microsandbox.dev
Cursor Anysphere
Supported
“// ~/.cursor/mcp.json { microsandbox: npx -y microsandbox-mcp }; Agent Skills works with Cursor.”
Source docs.microsandbox.dev
Codex CLI OpenAI
Supported
“Agent Skills 'Works with Claude Code, Cursor, Codex, Gemini CLI, GitHub Copilot ...'.”
Source docs.microsandbox.dev
Gemini CLI Google
Supported
“Agent Skills 'Works with ... Gemini CLI ...'.”
Source docs.microsandbox.dev
GitHub Copilot GitHub
Supported
“Agent Skills 'Works with ... GitHub Copilot'.”
Source docs.microsandbox.dev
Windsurf Codeium
Supported
“// ~/.codeium/windsurf/mcp_config.json { microsandbox: npx -y microsandbox-mcp }.”
Source docs.microsandbox.dev
OpenCode opencode
Supported
“// .opencode.json { microsandbox: { type: stdio, command: npx, args: [-y, microsandbox-mcp] } }.”
Source docs.microsandbox.dev
Zed Agent Zed
Supported
“Zed settings: context_servers.microsandbox command npx -y microsandbox-mcp.”
Source docs.microsandbox.dev
Pi earendil-works
Supported
“Run the Pi terminal coding agent against a project on your host.”
Source docs.microsandbox.dev

Plausible compatibility 10

Derived, not stored: the product passes SSH access, so any CLI coding agent is installable inside it. This is plausible compatibility, not a supported integration.

Aider
Amp
Antigravity
Devin
Droid
Goose
Hermes Agent
Kilo Code
Kiro CLI
Qwen Code

06 Compliance & regions Standards held, with ordinal rigor; never a score

Certifications held 0

No attestations found on the public surface. Empty means nothing is advertised, after checking, not that nothing was checked.

Regions —

Regions are not disclosed on the public surface.

07 Open source Apache-2.0 · fetched by machine, not quoted from marketing

Licence & repository

Licence: Apache-2.0OSI
Primary language: Rust
Stars: 8,666
Forks: 474
Created: 3 Oct 2024
Archived: No

Activity

Last commit: 12 Oct 2026
Latest release: v0.7.8 · 9 Oct 2026
Issues open / closed: 63 / 309 · 83% closed

Eight recent v0.7.x releases from September 16 to October 9, 2026; active pre-1.0 iteration, with explicit migration and breaking-change guidance.

Fetched via gh on 12 Oct 2026.

Self-hosting

Methods: package-managerbinarysource-build
You operate: —

Free under Apache-2.0 plus your own infrastructure. CLI and five SDKs embed/start the runtime without a setup server or long-running daemon. Linux requires KVM; macOS requires Apple Silicon; native Windows requires WHP (not WSL-only). Beta software may break APIs. Self-hosting lacks vendor-operated fleets, org billing/SSO/audit infrastructure; local supports more capabilities than cloud: live-memory checkpoints/forks, custom networking, historical logs, resource metrics, and local host mounts. Fully disconnected deployments are scoped in an evaluation; vendor makes no blanket air-gap claim. The vendor says this is the exact runtime it operates, not a separate proprietary core.

Documentation signals 8/8

SupportedQuickstart
SupportedAPI reference
SupportedSelf-host guide
SupportedChangelog
SupportedOpenAPI
Supportedllms.txt
SupportedExamples
SupportedArchitecture doc

08 Company & presence Facts only, not scored

Company

Headquarters: Unknown — not disclosed on scanned pages.
Founded: Unknown; repository created 2024-10-03.
Team size: Not disclosed
Hiring: Unknown — no working careers page found.

Activity

Last release: v0.7.8 on 2026-10-09T16:58:48Z github.com
Last blog post: 2026-06-01: Protect API keys in AI agent sandboxes. microsandbox.dev

Presence 6

Links found on the public surface scanned; not an exhaustive search.

09 Sources & method How this entry was produced

Public surface only

Everything here was read from microsandbox.dev's unauthenticated pages, docs, pricing and repository on 12 Oct 2026. Nothing behind a login was used.

Statuses, not scores

Capabilities are pass, partial or fail against a written rubric (features v21). “Not assessed” means the scan predates the key.

Confidence is separate

Each fact carries its own confidence. Low-confidence facts are shown, never hidden, so you can weigh them.

Schema: scan v5
Features registry: v21
Keys tracked today: 61
Assessed on this entry: 61

Something wrong on this page?

Nearby in the atlas

All Sandbox products →