// SCAN REPORT

Sandy

6 tools. 5 can modify or destroy data without limits.

1 destructive tool with no built-in limits. Policy required.

Last updated: 29 Jun 2026

5 can modify or destroy data

1 read-only

6 tools total

Community server · catalogue entry verified 29/06/2026

How to control Sandy ↓

TOOL MAP

What Sandy exposes to your agents

Read (1) Write / Execute (4) Destructive / Financial (1)

Running more than one server? Check your whole stack →

MOST DANGEROUS TOOLS

Critical Risk

The most dangerous Sandy tools

Destructive · High sandy_image While image creation is a Write operation, the tool's ability to delete the sandbox image makes this Destructive. Execute · High sandy_run The tool executes arbitrary TypeScript scripts, which is a code execution operation. Execute · Medium sandy_resume_session This tool resumes an execution session within a sandboxed TypeScript/AWS SDK environment. Execute · Low sandy_check The tool runs an operation (health check) in the sandboxed environment, which constitutes execution rather than a pure read.

5 of Sandy's 6 tools can modify, destroy, or commit something on every call — and an agent calls them with no built-in limits.

POLICY

How to control Sandy

PolicyLayer is an MCP gateway — it sits between your AI agents and Sandy, and nothing reaches the server without passing your rules. These are the rules we recommend:

Deny destructive operations

{
  "sandy_image": {
    "deny_if": [
      {
        "conditions": [],
        "on_deny": "Blocked by default. Requires approval."
      }
    ]
  }
}

Destructive tools should never be available to autonomous agents without human approval.

Rate limit write operations

{
  "sandy_create_session": {
    "limits": [
      {
        "counter": "sandy_create_session_per_hour",
        "window": "hour",
        "max": 30,
        "scope": "grant"
      }
    ]
  }
}

Prevents bulk unintended modifications from agents caught in loops.

Cap read operations

{
  "prime": {
    "limits": [
      {
        "counter": "prime_per_minute",
        "window": "minute",
        "max": 60,
        "scope": "grant"
      }
    ]
  }
}

Controls API costs and prevents retry loops from exhausting upstream rate limits.

Create a free account and register Sandy — nothing to install.
Add these rules — paste them, or build them visually. Tune the limits to your setup.
Point your MCP client (Claude, Cursor, anything) at your gateway URL.

ENFORCE POLICY ON SANDY →

Instant setup, no code required.

FULL CATALOGUE

All 6 Sandy tools

DESTRUCTIVE 1 tools

Destructive sandy_image Create or delete the Sandy sandbox image

EXECUTE 3 tools

Execute sandy_check Run a health check (baseline or connect). Uses an ephemeral session. Execute sandy_resume_session Resume an existing session and return its scripts path Execute sandy_run Run a TypeScript script in the Sandy sandbox

WRITE 1 tools

Write sandy_create_session Create a session and return its scripts path

READ 1 tools

Read prime Return the MCP SKILL.md content

EXPLORE

Related servers

Other MCP servers with similar tools — same risk classification, starter policies for each.

FAQ

Questions about Sandy

Can an AI agent delete data through the Sandy MCP server? +

Yes. The Sandy server exposes 1 destructive tools including sandy_image. These permanently remove resources with no undo. PolicyLayer blocks destructive tools by default so they never reach the upstream server.

How do I prevent bulk modifications through Sandy? +

The Sandy server has 1 write tools including sandy_create_session. Set a rate limit in your policy -- for example, 10 calls per hour prevents an agent from making more than 10 modifications per hour. PolicyLayer enforces this at the gateway, before calls reach Sandy.

How many tools does the Sandy MCP server expose? +

6 tools across 4 categories: Destructive, Execute, Read, Write. 1 are read-only. 5 can modify, create, or delete data.

How do I enforce a policy on Sandy? +

Register the Sandy MCP server in PolicyLayer, apply the suggested rules above (adjust the limits to your use case), and point your AI client at the PolicyLayer proxy URL instead of the server directly. Your agents keep the same tools; PolicyLayer evaluates every call against policy before it executes. Nothing to install, live in minutes.

Enforce policy on every Sandy tool call.

Deterministic rules across all 6 Sandy tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.

ENFORCE POLICY ON SANDY →

Instant setup, no code required.

6 Sandy tools catalogued and risk-classified — across an index of 43,000+ MCP servers.

// WHERE THIS COMES FROM

These policies come from Sandy's registry record.

The record behind this page: verified identity, auth posture, risk grade, every tool classified, recommended policy — re-checked continuously.

Teams ship this data inside their own products. See what a licence covers →