mxc
AGENTS.md2.9k starsChanged 12 days agoMITView on GitHub
Read and follow .github/copilot-instructions.md before writing or reviewing code. It contains MXC's architecture, compatibility, security, build, and SDK rules. The expectations below apply in addition. - Understand the existing flow before editing: callers, shared helpers, backend ownership, errors, cleanup, and affected SDK entry points. A working happy path is not enough; check how the change composes with existing behavior. - Put behavior where it belongs. Parsers decode contracts and preserve field presence; backend-specific defaults, policy checks, and enforcement belong in…
We split this AGENTS.md into the rules it gives an agent
11 rules. About 0 are common, 11 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈11
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 3
Read and follow .github/copilot-instructions.md before writing or reviewing code.
- Distinctive: line 13
Put behavior where it belongs.
- Distinctive: line 19
If the design cannot express the feature cleanly, refactor the responsible component while preserving existing callers and behavior.
- Distinctive: line 26
Consider whether a maintainer can understand and safely extend the result.
- Distinctive: line 30
If adding a feature requires replacing or deleting a large part of a subsystem, split the work into reviewable vertical slices: a small …
- Distinctive: line 38
Use plain, precise language in code reviews, comments, documentation, and handoffs.
- Distinctive: line 43
Keep comments sparse and short: explain a non-obvious reason or constraint, not what the next line does.
- Distinctive: line 48
Update the relevant canonical document rather than adding overlapping Markdown.
- Distinctive: line 49
Document only what the reader needs: supported behavior, usage, prerequisites, limitations, and useful error remedies.
- Distinctive: line 66
Prefer tests of useful behavior, feature interactions, and failure handling over many tiny checks of implementation details.
- Distinctive: line 74
Follow the validation runbook and backend guide.
≈ “Common” is an estimate: at least 1 in 5 of the 36,368 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
MXC agent instructions
Read and follow .github/copilot-instructions.md before writing or reviewing code. It contains MXC's architecture, compatibility, security, build, and SDK rules. The expectations below apply in addition.
Design for the codebase, not just the feature
- Understand the existing flow before editing: callers, shared helpers, backend ownership, errors, cleanup, and affected SDK entry points. A working happy path is not enough; check how the change composes with existing behavior.
- Put behavior where it belongs. Parsers decode contracts and preserve field presence; backend-specific defaults, policy checks, and enforcement belong in the backend, not shared parsing code.
- Reuse the normal implementation path. Do not add a parallel helper, a feature-specific variant, or a narrow exception just to make a feature work. If the design cannot express the feature cleanly, refactor the responsible component while preserving existing callers and behavior.
- Choose the simplest design that fits the current requirements. Add an abstraction only when it solves a real problem; do not build for imagined reuse. Name functions for their responsibility, not a growing list of cases.
- Consider whether a maintainer can understand and safely extend the result. Completing the requested feature does not excuse duplication, misplaced responsibilities, weakened checks, or hidden failure paths.
- If adding a feature requires replacing or deleting a large part of a subsystem, split the work into reviewable vertical slices: a small part of the feature working through the relevant layers, with its tests. Separate prerequisite refactoring from behavior changes; keep each step buildable and preserve existing behavior. Do not attempt the whole rewrite at once.
Write for everyday developers
Use plain, precise language in code reviews, comments, documentation, and handoffs. Explain the concrete behavior and consequence; avoid invented jargon and research-paper language. Be concise without leaving out the detail needed to understand or act. Neither verbosity nor oversimplification is helpful.
Keep comments sparse and short: explain a non-obvious reason or constraint, not what the next line does. Prefer clear names and straightforward code.
Documentation should describe the finished design, behavior, and usage, not an agent's thought process, implementation diary, or discarded approaches. Update the relevant canonical document rather than adding overlapping Markdown. Document only what the reader needs: supported behavior, usage, prerequisites, limitations, and useful error remedies. Include implementation detail only when it explains a necessary maintenance or design decision, in developer docs.
Separate paragraph-length Markdown list items with a blank line; keep short lists compact.
Require meaningful behavior coverage
- For a feature, add or extend a functional validation test in the applicable CI suite. Backend behavior must run in scheduled/nightly validation on the relevant hosts; adding a test file without wiring it into CI is not enough.
- For a bug fix, add a regression scenario that exercises the reported failure through the relevant public API or backend flow and checks the correct result. Where practical, show that it fails before the fix and passes afterward.
- Prefer tests of useful behavior, feature interactions, and failure handling over many tiny checks of implementation details. Extend existing scenarios rather than duplicating them.
- Unit tests, including tests at the bottom of
.rsfiles, are useful for isolated logic but do not replace functional validation. Do not judge coverage by test count or tests that merely repeat the new implementation. - Follow the validation runbook and backend guide. Check that the intended scenario actually runs and asserts behavior on a suitable host. State what was not run or cannot be covered; do not invent evidence. Documentation-only changes do not need runtime tests.
For reviews, use the MXC review skill. Focus on high- and medium-severity defects and consolidate findings into one review, rather than creating extra review rounds for nits or optional cleanup.
Trending
Steady: 0 new projects took it in the last 4 weeks, none the 4 weeks before.thin data
Who's talking about microsoft/mxc
- A Hacker News story“MXC - a sandboxed code execution system”Hacker News · Oct 9, 2026 · 196 points
- @sungkim.bsky.social“Surprisingly, this is cross-platform. Microsoft eXecution Container (MXC) MXC is a sandboxed code execution …”Bluesky · Jun 6, 2026 · 85 likes
- @five-oclick-tech.bsky.social“MXC: Microsoft's sandbox for untrusted code Microsoft released MXC, a cross-platform sandboxed execution …”Bluesky · Oct 10, 2026 · 5 likes
- golang/go“syscall: update Windows SysProcAttr for recent process creation features”GitHub · Jul 15, 2026 · 4 reactions
- A Hacker News story“Microsoft Execution Containers (MXC)”Hacker News · Jun 2, 2026 · 3 points
The biggest 5 of 66 mentions linking to the repository that our crawl found. Each line is their own headline.
See all 66
Talked about on GitHub, the web, Bluesky and Hacker News (66 times).
On GitHub · 39
- Jul 15, 2026syscall: update Windows SysProcAttr for recent process creation features — golang/go4 reactions · 32 comments
- Oct 19, 2026improve: run isolated sessions through the official MXC SDK — openclaw/openclaw-windows-packaging1 reaction · 11 comments
- Oct 4, 2026feat(node): honor the typed system.run execution context — openclaw/openclaw-windows-node1 reaction · 4 comments
- Jul 7, 2026System requirement doc for different Platform of sandboxing feature — github/copilot-cli1 reaction
- Jun 26, 2026feat(mxc): add Windows MXC sandbox backend — openclaw/openclaw1 reaction · 47 comments
34 more on GitHub
- Oct 29, 2026Explainer: Tool Calls: How LLMs Act in the World — pank2015/signalgraph-site0 reactions · 1 comment
- Oct 22, 2026Codex 0.162.1 MXC rejects an explicit protected deny path before launching a synthetic payload — openai/codex0 reactions
- Oct 22, 2026[Feature]: Add Microsoft MXC sandbox backend support — zeroclaw-labs/zeroclaw0 reactions
- Oct 21, 2026chore(deps): bump the cargo-minor-and-patch group with 8 updates — everruns/yolop0 reactions · 1 comment
- Oct 19, 2026NVX Phase 3: experimental streaming and persistent lifecycle — huzaifa-d/mxc0 reactions
- Oct 19, 2026NVX Phase 1: Minimal Signed MicroVM integration — huzaifa-d/mxc0 reactions
- Oct 19, 2026Bump the nuget-remaining group with 27 updates — w0rldx/XE-Local-AI-Engine.Source0 reactions · 2 comments
- Oct 17, 2026Bump Microsoft.Mxc.Sdk from 1.0.0 to 1.0.1 — MGrunwald42/safecraft-mcp0 reactions
- Oct 16, 2026AD-1334: an MXC Tier-2 isolation backend - kernel-enforced sandboxing behind the AD-993 protocol (revives AD-995) — seangalliher/ProbOS0 reactions · 1 comment
- Oct 14, 2026chore(mxc): move the Linux executor to MXC v1.0.0 — ROCm/axis0 reactions · 1 comment
- Oct 11, 2026feature request: sandbox game execution (with MXC?) — Pathoschild/SMAPI0 reactions · 3 comments
- Oct 9, 2026HN Digest - 2026-10-09 — Emerson4892/agentic-workflows-workshop0 reactions
- Oct 9, 2026HN Digest - 2026-10-09 — Rodro-31/agentic-workflows-workshop0 reactions
- Oct 9, 2026HN Digest - 2026-10-09 — SandroZ115/agentic-workflows-workshop0 reactions
- Oct 9, 2026HN Digest - 2026-10-09 — vgarcial10/agentic-workflows-workshop0 reactions
- Oct 9, 2026[RESEARCH] Factor MXC + Gemini agent architecture into Trust/Control and Reach/Motion — bdf1992/Soveraeign0 reactions
- Oct 9, 2026mxc : Policy-driven, layered isolation and containment, 关注这种分层思想,和设计中采用的分层思想类似 — coolplayagent/agent-computer0 reactions
- Oct 8, 2026GitHub Changelog:Local sandboxing for GitHub Copilot now generally available — forks-felickz/github-changelog-reader0 reactions
- Oct 8, 2026GitHub Changelog:Local sandboxing for GitHub Copilot now generally available — renefritze/github-changelog-reader0 reactions
- Oct 8, 2026Sandbox scripts — ekolis/FrEee0 reactions
- Oct 8, 2026GitHub Changelog:Local sandboxing for GitHub Copilot now generally available — abirismyname/github-changelog-reader0 reactions
- Oct 8, 2026Seek 2026-10-08 08:23 IDT: 17 Habitat desk-bot news posts — michelabboud/aitamer-news0 reactions · 2 comments
- Oct 8, 2026content: prepare Daily AI Pulse for 2026-10-08 — GadDev/daily-ai-pulse0 reactions
- Oct 2, 2026Spike: qualify MXC Hyperlight rich Python continuity and durable recovery — sokolaidev/maf-extensions0 reactions · 4 comments
- Oct 2, 2026Integrate rich Python through MXC Hyperlight with optional durable sessions — sokolaidev/maf-extensions0 reactions
- Sep 14, 2026[Draft prototype] Known-tool policy catalog (Policy Store design) — ChazGo/mxc0 reactions
- Sep 4, 2026Bundle mxc for Windows — openkaiden/kaiden0 reactions · 3 comments
- Aug 12, 2026Decide production agent sandbox framework: AppContainer, OpenClaw Docker, or MXC — microsofthackathons/MicroClaw0 reactions · 1 comment
- Aug 5, 2026Can't create Sandbox on Windows when BaseContainer is velocity-gated and bfscfg.exe is unavailable — microsoft/vscode0 reactions · 4 comments
- Jul 16, 2026Experimenting with AppContainers — hugsy/pwn--0 reactions
- Jun 18, 2026feat: MXC sandbox integration for contained local code execution — microsoft/entrabot0 reactions · 1 comment
- Jun 8, 2026docs: MXC (Microsoft Execution Containers) platform research — microsoft/entrabot0 reactions
- Jun 6, 2026Research MXC as opt-in sandbox backend for scoped shell execution; evaluate MCP enforcement separately — ZaxbyHub/opencode-swarm0 reactions · 3 comments
- May 21, 2026Enable sandboxing in windows using MXC SDK — microsoft/vscode0 reactions · 1 comment
On the web · 14
- Oct 14, 2026MXC - Microsoft's sandbox for code you shouldn't trust! · Niels Tanis — blog.fennec.dev
- Oct 13, 2026Microsoft Execution Containers 2026: Sandbox Your AI Agents — The Vibelog — thevibelog.dev
- Oct 12, 2026Microsoft Makes Execution Containers Generally Available to Sandbox AI Agents — dev.to
- Oct 9, 2026Microsoft MXC - AI Agent Review — agentconn.com
- Oct 7, 2026Microsoft Execution Containers: Policy-driven ... - Windows Blog — blogs.windows.com
9 more on the web
- Sep 24, 2026Your AI agent is one command from disaster. — linkedin.com
- Sep 12, 2026CTO at NCSC Summary: week ending September 13th — ctoatncsc.substack.com
- Jun 29, 2026Inside Microsoft Execution Containers: How Windows ... - itnext — medium.com
- Jun 16, 2026MXC、OpenClaw、Microsoft Scoutの違いとは?WindowsでAIエージェントを安全に動かす仕組みを整理する #MicrosoftBuild — qiita.com
- Jun 12, 2026ハンズオン!夢が詰まったサンドボックスシステム、Microsoft Execution Containers(MXC)とは #MSBuild — qiita.com
- Jun 5, 2026微软开源MXC:给你的AI Agent套上安全沙箱,3分钟上手配置_微软 mxc-CSDN博客 — blog.csdn.net
- Jun 4, 2026当Windows成为Agent的监狱-操作系统级Agent安全架构深度解读-CSDN博客 — blog.csdn.net
- Jun 3, 2026Microsoft Build 2026: Top announcements from a DevOps ... — dev.to
- Jun 3, 2026Microsoft Execution Containers (MXC) User Training Course|platypus2000jp — note.com
On Bluesky · 10
- Jun 6, 2026Surprisingly, this is cross-platform. Microsoft eXecution Container (MXC) MXC is a sandboxed code execution system for running untrusted… — @sungkim.bsky.social85 likes and reposts · 5 replies
- Oct 10, 2026MXC: Microsoft's sandbox for untrusted code Microsoft released MXC, a cross-platform sandboxed execution system for untrusted code like AI… — @five-oclick-tech.bsky.social5 likes and reposts
- Oct 8, 2026🌐 github.com/microsoft/mxc blogs.windows.com/windowsdeve... — @michabbb.bsky.social2 likes and reposts
- Aug 12, 2026Microsoft eXecution Containers (MXC) are an opensource, cross platform execution sandbox for running untrusted code (esp. AI agents). Still… — @arsalan-zaidi.mastodon.social.ap.brid.gy2 likes and reposts
- Oct 3, 2026Cooooooomplicated af to say the least. We've been building out a ton of functionality on this front through MXC as well as Windows… — @syfuhs.net1 likes and repost
5 more on Bluesky
- Jun 6, 2026📦 microsoft / mxc ⭐ 469 (+57) 🗒 Rust Policy-driven, layered isolation and containment — @github-trending.bsky.social1 likes and repost
- Jun 5, 2026WindowsやLinuxなどで信頼できないコードを安全に動かせるサンドボックス実行環境らしい マルチプラットフォーム対応でセキュリティ強化に役立ちそうかも yug1224 starred microsoft/mxc… — @yug1224.com1 likes and repost
- Oct 10, 2026MXCのバックエンドにWSLC使えばLinuxコマンドも使える環境をセットアップできるってことか。Windowsの環境これでいいじゃん github.com/microsoft/mxc — @soisoiro.bsky.social0 likes and reposts · 1 reply
- Oct 8, 2026Microsoft execution container blogs.windows.com/windowsdevel... Et son repo github github.com/microsoft/mxc — @aymericw.bsky.social0 likes and reposts · 1 reply
- Jun 8, 2026🚀 Skyrocketing! 🚀 (200+ new stars) 📦 microsoft / mxc ⭐ 753 (+211) 🗒 Rust Policy-driven, layered isolation and containment — @github-trending-rs.bsky.social0 likes and reposts · 1 reply
On Hacker News · 3
- Oct 9, 2026MXC - a sandboxed code execution system196 points · 85 comments
- Jun 2, 2026Microsoft Execution Containers (MXC)3 points
- Jun 4, 2026Microsoft EXecution Container (MXC)2 points
Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.
Of the AGENTS.md files we read, the 28,693 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.
In this repository3 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- mxc-code-reviewSkill
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

