Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/d3v1an/ssh-mcp/claude-mdgit clone --depth 1 https://github.com/d3v1an/ssh-mcpWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/instructions/d3v1an/ssh-mcp/claude-md)<a href="https://agentmods.dev/instructions/d3v1an/ssh-mcp/claude-md"><img src="https://agentmods.dev/badge/instructions/d3v1an/ssh-mcp/claude-md.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.01551 | $0.01551 |
| Opus 5 | $0.00776 | $0.00776 |
| Sonnet 5 | $0.00310 | $0.00310 |
| Haiku 4.5 | $0.00155 | $0.00155 |
Grade A, and why
ssh-mcp CLAUDE.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 62 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CLAUDE.md
This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
Project Overview
MCP (Model Context Protocol) server for SSH remote server administration. Enables AI models to manage remote servers over SSH — executing commands, transferring files via SFTP, and reading/writing remote files. Documentation is in Spanish.
Commands
npm run build # Compile TypeScript → dist/
npm start # Run MCP server (dist/index.js)
No test or lint commands are configured yet.
Architecture
Single-class MCP server (SSHMCPServer in src/index.ts) that manages one SSH connection at a time via the ssh2 library.
Source Files
src/index.ts— Main server class. Handles MCP tool dispatch, SSH/SFTP connection lifecycle, command execution, interactive exec (PTY + auto-response), persistent shell sessions, and file operations. All tool handlers are methods onSSHMCPServer. Version is read dynamically frompackage.json.src/tools.ts— MCP tool definitions (name, description, JSON Schema). 17 tools:ssh_list_profiles,ssh_connect,ssh_disconnect,ssh_status,ssh_exec,ssh_exec_interactive,ssh_shell_start,ssh_shell_send(includessensitiveflag),ssh_shell_read,ssh_shell_close,ssh_upload,ssh_download,ssh_ls,ssh_read_file,ssh_write_file,ssh_history,ssh_undo.src/profiles.ts— Loads and validates SSH profiles fromprofiles.jsonat startup (not on first use). Validates required fields (host,port,username,privateKeyPath,hostFingerprint), verifies the key file is readable, and caches private keys in memory. Never exposes private keys or passphrases in responses.src/security.ts— Dangerous command detection (regex patterns) andAuditLogger. UsesWriteStreamfor non-blocking writes with0o600permissions (enforced viachmodSyncon existing files). Common secret patterns (password=, token=, Bearer) are redacted before writing.src/types.ts— TypeScript interfaces:SSHProfile(includeshostFingerprint,localSandboxDir),AuditEntry,PromptResponse,ShellSession,CommandRecord,ReverseInfo.src/utils.ts— Pure utility functions:formatUptime,padRight,escapeShellArg,stripAnsi.src/validation.ts— Input validation helpers:requireString,optionalString,optionalBoolean,optionalNumber,clampTimeout(bounds: 1s–5min).
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 4d ago First seen · 62 lines · 1,551 tokens per session scan A 1989f02d589a
ssh-mcp CLAUDE.md is an instructions file published in the GitHub repository d3v1an/ssh-mcp (0 stars, last pushed 3mo ago), licensed MIT. It adds 1,551 tokens to every session, about $0.0078 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
awsmux AGENTS.md
Instructions for 0hardik1/awsmux, covering agents.md, what awsmux is, build, test, verify, layout and the safety model — invariants you must not weaken.
plan-forge naming.instructions.md
Azure naming conventions — Microsoft CAF abbreviations, constraints, uniqueness, tagging.
agent-skills fortrabbit.instructions.md
Instructions for fortrabbit/agent-skills, covering fortrabbit — github copilot instructions, what fortrabbit is, configuration lookup, available operations and mcp server (optional).
cron-mcp CLAUDE.md
Claude Code instructions for Shreyas-Profile/cron-mcp, covering deploying cron-mcp to hetzner + regiq.in, server (shared with the rest of the globalion mcp fleet), deploy a new version, first-time subdomain wiring (cron.regiq.in) and env vars.
Railway-mcp CLAUDE.md
Claude Code instructions for Travis-Gilbert/Railway-mcp, covering railway mcp server, quick reference, file map, architecture and key patterns.
equinix-fabric-mcp CLAUDE.md
Claude Code instructions for sliuuu/equinix-fabric-mcp, covering claude.md — equinix-fabric-mcp, canonical files, auth, add equinixclientid and equinixclientsecret and running.