Remote Agents MCP server
Control fleets of remote machines via AI agents: exec, files, git, fleet ops, map/reduce.
1 stars54 downloads/wk
Reviews
Write oneNobody has reviewed Remote Agents yet.
If you have run it, two minutes of your experience saves the next person an afternoon.
Remote Agents tools
No tool declarations could be read from the package source. They show once the server is installed.
Public scan report
scanner v0.1.9 · 2026-09-21 · same rubric, same numbers if you re-run it
- Code scan3 source files scanned15/25
- –Live reliabilityno gateway calls yet and no remote to proben/a
- –Tool poisoningtools not inspected (local package is not executed); not countedn/a
- Auth qualitylocal package, no credentials required12/15
- Maintenancelast push 28 days ago15/15
- Maintainer identityregistry namespace matches repository owner6/10
Findings (2)
- mediumnpm install lifecycle script present
install.script - mediumnpm install lifecycle script present
install.scriptpackage.json: …ents": "run.js" }, "scripts": { "postinstall": "node install.js" }, "files": [ …
What the publisher says
From the Remote Agents repository's README, as published. We do not edit it. Read it on GitHub
Remote Agents
A unified, MCP-compatible system for controlling fleets of remote machines through AI agents (Claude, opencode). Agents connect outbound to a relay; an MCP server lets the AI run commands, manage files, drive git, schedule tasks, and orchestrate the whole fleet — all over end-to-end-encrypted channels.
Features
MCP stdio server (mcp), or installs itself as a service (install).
- Single Rust binary (remote-agent) — runs as an agent daemon (run), an
ciphertext.
- End-to-end encryption (AES-GCM-256) on by default; the relay forwards only
bypass, disabled — with path/command allow- & deny-lists.
- Safety modes per host — plan (read-only), edit (writes with backups),
across all agents, by tags, or by OS family; results aggregated per host.
- Fleet as one computer — run any operation (exec/read/write/git)
command, reduce the outputs, with per-partition retry.
- Distributed MapReduce — partition data across the fleet, map with a shell
credentials (token-saving orchestration).
- Autonomous mode — delegate AI tasks to a host that runs them with its own
self-hosted Rust WebSocket relay; switch by changing relay_url.
- Two interchangeable relays — Cloudflare Workers (Durable Objects) or a
(sendfile) or rsync-like directory sync (syncdir), SHA-256 verified.
- Direct UDP data channel (QUIC) with hole-punching and WebSocket fallback.
- File & folder transfer host→host over that channel — single files
Architecture
┌──────────────────────────────────────────────────────────────┐
│ Any MCP client — Claude Code / Desktop, Cursor, Cline, Roo, │
│ Kilo, Windsurf, Zed, opencode, Continue, Goose │
│ remote-agent mcp (Rust binary, MCP stdio server) │
└───────────────────────────────┬──────────────────────────────┘
│ wss:// (control + UDP signaling)
▼
┌───────────────────────────────────┐
│ Relay (rooms route by token) │
│ CF Worker or self-hosted Rust │
└───────────────────────────────────┘
▲ ▲ ▲
│ wss │ wss │ wss
┌──────┴─────┐ ┌──────┴─────┐ ┌──────┴─────┐
│ Agent │ │ Agent │ │ Agent │
│ (daemon) │ │ (daemon) │ │ (daemon) │
└──────┬─────┘ └──────┬─────┘ └──────┬─────┘
└─────────────┐│┌─────────────┘
direct UDP / QUIC data channel
(hole-punched peer-to-peer; bulk file & folder
transfer; automatic relay fallback behind NATTwo planes: control (commands + results) and UDP signaling always go through the relay over wss:// (the relay sees only ciphertext); bulk data (sendfile / syncdir) rides a direct UDP/QUIC channel hole-punched between the two peers, falling back to the relay when NAT blocks the punch.
Workspace layout
Install
Shortened. The full README is on GitHub.
Nothing above is checked by us. What we check is on the safety report.
Install directly
Runs npx -y remote-agents on your machine. Read the scan report first; the gateway never runs local packages.
claude mcp add remote-agents -- npx -y remote-agents
Remote Agents: common questions
- Is Remote Agents MCP server safe?
- Mostly: it is graded B (74/100). Read the Remote Agents safety report
- How do I install Remote Agents?
- It runs on your machine. Copy the Claude Code, Claude Desktop or Cursor config from the install section.
- Does Remote Agents need an API key?
- Not as far as the registry entry and our scan can tell: no credentials are declared or required.
- Is Remote Agents maintained?
- The last commit was 29 days ago (2026-08-24). The latest release is v0.1.40.
- What can I use instead of Remote Agents?
- Servers from other publishers that do the same job: Ocm MCP server, Prumo MCP server and Tenki Sandbox MCP server. Compare all Remote Agents alternatives.
Alternatives to Remote Agents
Same job from other publishers: the closest match first, then the best rated.
- OcmGuardrailed fleet ops for AI agents: multi-cluster Kubernetes via OCM with policy, approval, audit.not reviewedGrowingB
- PrumoChecks the context files a coding agent reads against the git index, and reports what drifted.not reviewedGrowingA
- Tenki Sandbox MCPDisposable microVM sandboxes for AI agents: run code, read/write files, git, preview URLs.not reviewedGrowingA
- GitComprehensive Git MCP server enabling native git tools including clone, commit, worktree, & more.not reviewedEstablishedA
- NotionToken-auth Notion MCP: pages, databases, blocks, comments, files. Runs headless in CI and agents.not reviewedEstablishedB