shell

Execute shell command on remote client (cmd.exe on Windows, /bin/sh on Linux). If only one client connected, client_id is auto-selected.

SERVERClaude C2 SOURCE0xyg3n/claude-c2

High RISK CLASS

Category Execute

Parameters 00 required

Recommended Rate-limitedsee the rule below

Registry record Grade F, identity unverified Pull the record →

This record as markdown: /tools/0xyg3n-claude-c2/shell.md

WHAT IT DOES

What shell does on Claude C2

AI agents invoke shell to trigger actions in Claude C2. What it does depends on the arguments the agent supplies, and its effects often reach beyond the immediate call: builds kicked off, notifications sent, workflows started.

RISK

Why shell is rated High

This tool allows arbitrary shell command execution on remote systems with minimal constraints. Shell commands can read, write, delete files, exfiltrate data, install malware, pivot to other systems, and cause widespread damage depending on the client's privilege level. The blast radius is unlimited (depends entirely on what shell commands an AI agent chooses to run).

From the tool's definition Tool description explicitly states "Execute shell command on remote client (cmd.exe on Windows, /bin/sh on Linux)". The term "Execute" combined with unrestricted shell command capability on remote systems represents the highest-impact execution vector.

Attacks that exploit this kind of access

RECOMMENDED RULE

The rule that runs shell safely

PolicyLayer is an MCP gateway: it sits between your AI agents and Claude C2, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For shell, this is the rule to start with:

shell Rate-limited

shell stays usable, but rate-capped: a runaway agent can't fire it dozens of times a minute. Everything else on the server is denied unless you say otherwise.

View as policy code

policy.json

{
  "version": "1",
  "default": "deny",
  "tools": {
    "shell": {
      "limits": [
        {
          "counter": "shell_rate",
          "window": "minute",
          "max": 10,
          "scope": "grant"
        }
      ]
    }
  }
}

RATE-LIMIT THIS TOOL → Instant setup, no code required.

The button opens the PolicyLayer dashboard: create your workspace, connect Claude C2, apply this rule, and every shell call is checked against it from then on.

FAQ

Questions about shell

What does the shell tool do? +

Execute shell command on remote client (cmd.exe on Windows, /bin/sh on Linux). If only one client connected, client_id is auto-selected. It is categorised as a Execute tool in the Claude C2 MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.

How do I enforce a policy on shell? +

Register the Claude C2 MCP server in PolicyLayer and add a rule for shell: allow, deny, rate-limit, or require approval. Point your MCP client at the PolicyLayer proxy URL and the rule is enforced on every call, before it reaches Claude C2. Nothing to install.

What risk level is shell? +

shell is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.

Can I rate-limit shell? +

Yes. Add a rate_limit block to the shell rule in your PolicyLayer policy. For example, setting max: 10 and window: 60 limits the tool to 10 calls per minute. Rate limits are tracked per agent session and reset automatically.

How do I block shell completely? +

Set action: deny in the PolicyLayer policy for shell. The AI agent will receive a policy violation error and cannot call the tool. You can also include a reason field to explain why the tool is blocked.

What MCP server provides shell? +

shell is provided by the Claude C2 MCP server (0xyg3n/claude-c2). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.

KEEP EXPLORING

More on Claude C2, and thousands of servers like it.

This server

All 94 Claude C2 tools → High-risk tools on this server →

Across the catalogue

adb_shell_command on Android Forensics ADB MCP Server → All high-risk MCP tools → The MCP Attack Database →

Guides

Govern code and CI agents →Agent sandbox →Least privilege for MCP →