mcp-ssh

mcp-ssh is an MCP (Model Context Protocol) server that bridges AI clients to SSH functionality. It lets an AI assistant create/manage SSH connections, execute commands (including composite commands), manage tmux sessions, and perform file and process operations on remote servers using structured MCP tool calls.

Evaluated Mar 30, 2026 (22d ago)
Repo ↗ DevTools mcp ssh tmux devops automation terminal nodejs typescript remote-management
⚙ Agent Friendliness
47
/ 100
Can an agent use this?
🔒 Security
40
/ 100
Is it safe for agents?
⚡ Reliability
28
/ 100
Does it work consistently?

Score Breakdown

⚙ Agent Friendliness

MCP Quality
70
Documentation
60
Error Messages
0
Auth Simplicity
40
Rate Limits
0

🔒 Security

TLS Enforcement
40
Auth Strength
45
Scope Granularity
20
Dep. Hygiene
55
Secret Handling
45

Strengths: supports SSH key and password auth to remote hosts; README mentions timeouts and structured request validation via Zod (claims). Risks/unknowns: no described auth controls for the MCP server itself, no RBAC/scopes, and no stated audit/log redaction strategy. Rate limiting and explicit command allowlisting/sandboxing are not documented. TLS for MCP transport and secret handling behavior (e.g., logging) are not specified in the provided README/manifest content.

⚡ Reliability

Uptime/SLA
0
Version Stability
50
Breaking Changes
0
Error Recovery
60
AF Security Reliability

Best When

You already use an MCP-capable client (e.g., Cursor) and want an AI-driven interface for controlled SSH/tmux operations in a developer/ops workflow.

Avoid When

You need standardized web APIs, strong enterprise security controls (SSO/RBAC), or you cannot enforce policies around what commands/files the AI may access.

Use Cases

  • Remote command execution on servers via an AI assistant
  • Persistent remote workflows using tmux session management
  • File upload/download and viewing remote files through MCP tools
  • Automation of common DevOps tasks (process monitoring, command retries/timeouts)
  • Reducing prompt-to-shell-command friction for operators

Not For

  • Replacing a fully managed remote execution service with robust access controls and auditing
  • Handling untrusted, arbitrary command input without strict sandboxing/policy
  • Use cases requiring a documented public REST/GraphQL/SDK interface beyond MCP
  • High-assurance environments that require formally specified security guarantees

Interface

REST API
No
GraphQL
No
gRPC
No
MCP Server
Yes
SDK
No
Webhooks
No

Authentication

Methods: SSH password authentication (as described) SSH key authentication (as described)
OAuth: No Scopes: No

Authentication is for SSH to remote hosts (password/key) rather than API-style auth for mcp-ssh itself. README does not describe any MCP server auth, scopes, or tenant isolation.

Pricing

Free tier: No
Requires CC: No

Open-source npm package; cost is self-hosting/operational overhead.

Agent Metadata

Pagination
none
Idempotent
False
Retry Guidance
Not documented

Known Gotchas

  • Operations can run long or block; README emphasizes blocking detection and waiting behavior.
  • AI-driven command execution requires policy to avoid destructive or sensitive commands.
  • tmux collaboration guidance suggests creating tmux and waiting for command completion; agents may otherwise interleave commands incorrectly.
  • If required project files are removed (README warns not to delete the cloned directory), the MCP bridge may fail.

Alternatives

Full Evaluation Report

Comprehensive deep-dive: security analysis, reliability audit, agent experience review, cost modeling, competitive positioning, and improvement roadmap for mcp-ssh.

AI-powered analysis · PDF + markdown · Delivered within 30 minutes

$99

Package Brief

Quick verdict, integration guide, cost projections, gotchas with workarounds, and alternatives comparison.

Delivered within 10 minutes

$3

Score Monitoring

Get alerted when this package's AF, security, or reliability scores change significantly. Stay ahead of regressions.

Continuous monitoring

$3/mo

Scores are editorial opinions as of 2026-03-30.

8642
Packages Evaluated
17761
Need Evaluation
586
Need Re-evaluation
Community Powered