Skip to content

CONTRIBUTING.md

Latest commit

 

History

History
119 lines (84 loc) · 5.13 KB

CONTRIBUTING.md

File metadata and controls

119 lines (84 loc) · 5.13 KB
Raw

Contributing to Awesome AI Sandboxes

This list aims to be the canonical factual reference for AI agent sandbox providers. Consistency and source rigor matter more than volume.

Scope

In scope:

  • Code execution sandboxes (microVMs, containers, gVisor, etc.)
  • Browser sandboxes for AI agents
  • Desktop / computer-use sandboxes
  • Self-hostable sandbox orchestrators

Out of scope:

  • Model hosts (Anthropic, OpenAI, AWS Bedrock, etc.)
  • Pure GPU rental without sandbox/isolation positioning (vast.ai, RunPod, Lambda Labs)
  • Orchestration-only tools without sandboxing (Composio, AgentOps, LangSmith)
  • IDE/editor sandboxes positioned at end users, not agents

If you're unsure, open an issue before drafting a PR.

Source rules

  1. All claims must come from official sources only — the provider's docs, landing page, or official GitHub repo.
  2. No blog posts, third-party articles, tweets, or marketing collateral as primary sources.
  3. Keep entries factual — no marketing language, no superlatives, no comparisons to other providers.

Examples:

  • ✅ "Sub-200ms cold starts" sourced from e2b.dev/docs
  • ✅ "Apache 2.0" sourced from the GitHub LICENSE file
  • ❌ "The fastest sandbox provider"
  • ❌ "TechCrunch reported $50M Series A"

Section placement

License Section
Apache 2.0, MIT, AGPL, BSD, MPL (any OSI-approved license) Open Source
Proprietary, closed-source, source-available with restrictions Closed Source

Within each section, append new entries at the bottom.

Entry template

Every entry needs the 8 required fields in this exact order. The 3 fields marked (recommended) should be included if the data is in the provider's docs.

### [Name](https://provider.dev)
[Website](https://provider.dev) | [Docs](https://docs.provider.dev) | [GitHub](https://github.com/org/repo)

One- or two-sentence factual description. ≤ 25 words. Should tie to the sandbox theme.

- **Isolation:** {tech}
- **Key features:** {comma-separated list, 5–8 distinctive items}
- **Cold start:** {recommended — e.g., "<200ms", "Sub-90ms", "~25ms", or "Not specified"}
- **Max session:** {recommended — e.g., "24h", "Unlimited", "5h", or "Not specified"}
- **Snapshots / Forking:** {recommended — "Yes (snapshot + fork)" / "Yes (snapshot only)" / "No" / "Not specified"}
- **Stateful:** {Yes / Yes (with volumes) / Ephemeral / No / Not specified}
- **GPU:** {Yes (model list) / Yes / No / Not specified}
- **BYOC / Self-host:** {Yes / Yes (Enterprise) / Yes (deploy-to-VPC) / No / Not specified}
- **SDKs:** {languages, with package names in backticks}
- **License:** {Apache 2.0 / MIT / AGPL 3.0 / etc. — required for Open Source section}
- **Pricing:** {free tier description; paid model; starting price}

---

Standardized values

To keep entries comparable, use these enums where applicable.

Stateful: Yes / Yes (with volumes) / Ephemeral / No / Not specified

GPU: Yes (<model list>) / Yes / No / Not specified

BYOC / Self-host: Yes / Yes (Enterprise) / Yes (deploy-to-VPC) / No / Not specified

Snapshots / Forking: Yes (snapshot + fork) / Yes (snapshot only) / No / Not specified

Description guidelines

  • ≤ 25 words, 1–2 sentences
  • Ties to "sandbox" or related vocabulary
  • States what the product IS, not what it's better than

Good: "Open-source browser API for AI agents and apps. Sandboxed Chrome sessions with anti-detection, session viewer, and computer-use integrations."

Bad: "The fastest, most secure browser sandbox ever built."

Submission checklist

  • Entry placed in correct section (license-based)
  • Field order matches template exactly
  • All 8 required fields present
  • All 3 recommended fields included if data is available
  • All claims link to official sources
  • Description ≤ 25 words and ties to the sandbox theme
  • Trailing --- separator included
  • PR title: Add <Name>
  • PR body links to the official sources used

Review criteria

PRs are reviewed for:

  1. Source verification — every factual claim traces to an official source
  2. Template compliance — field order, formatting, section placement
  3. Scope — in/out per the scope section above
  4. Factual neutrality — no marketing language, no comparative claims

Field design notes

The following attributes were considered as top-level fields and deliberately kept inside Key features as free text instead. Future contributors are welcome to propose promoting them via a separate issue if the case strengthens.

Considered Rationale for not promoting to a top-level field
Compliance / certifications (SOC 2, HIPAA, ISO 27001) Mostly negotiated per-contract, rarely a public-tier feature. Remains valid inside Key features when officially advertised.
Max concurrency Overlaps with provider-specific scale claims already captured in Key features (50k+ sandboxes, millions of VMs, etc.).
Resources per sandbox (CPU / RAM / disk) Varies by plan tier; not a fixed comparable.
Computer use / desktop support Binary field would miss nuance (Linux-only vs cross-OS vs browser-only). Stays in Key features as text.