// SCAN REPORT

AgentBridge MCP Server

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

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

Last updated: 2 Jul 2026

5 can modify or destroy data

7 read-only

12 tools total

Community server · catalogue entry verified 02/07/2026

How to control AgentBridge MCP Server ↓

TOOL MAP

What AgentBridge MCP Server exposes to your agents

Read (7) Write / Execute (4) Destructive / Financial (1)

Running more than one server? Check your whole stack →

MOST DANGEROUS TOOLS

Critical Risk

The most dangerous AgentBridge MCP Server tools

Financial · High execute_refund_order Although the description notes the refund is SIMULATED and no real payment processor is called, the tool's core purpose is to execute financial refund operations. Execute · High call_action This is a dynamic action invocation tool that can trigger arbitrary actions from a manifest. Write · Medium draft_refund_order The tool creates a draft record (a write operation) but does not execute or finalize the refund. Write · Medium add_internal_note This tool creates new data (an internal note) and modifies the order record by appending metadata.

5 of AgentBridge MCP Server's 12 tools can modify, destroy, or commit something on every call — and an agent calls them with no built-in limits.

POLICY

How to control AgentBridge MCP Server

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

Block financial tools by default

{
  "execute_refund_order": {
    "deny_if": [
      {
        "conditions": [],
        "on_deny": "Requires human approval."
      }
    ]
  }
}

Financial tools should be explicitly enabled per use case, not open by default.

Rate limit write operations

{
  "add_internal_note": {
    "limits": [
      {
        "counter": "add_internal_note_per_hour",
        "window": "hour",
        "max": 30,
        "scope": "grant"
      }
    ]
  }
}

Prevents bulk unintended modifications from agents caught in loops.

Cap read operations

{
  "discover_manifest": {
    "limits": [
      {
        "counter": "discover_manifest_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 AgentBridge MCP Server — 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 AGENTBRIDGE →

Instant setup, no code required.

FULL CATALOGUE

All 12 AgentBridge MCP Server tools

FINANCIAL 1 tools

Financial execute_refund_order Executes a previously drafted refund. SIMULATED — no real payment processor is called. Marks the order as refu

EXECUTE 1 tools

Execute call_action Invoke an AgentBridge action. Risky actions return a confirmationRequired response with a confirmationToken; t

WRITE 3 tools

Write add_internal_note Adds an internal note to an order. Visible to staff only. Write create_thing Creates a new thing. Replace with a real action from your app. Write draft_refund_order Creates a refund draft for an order. The draft is not executed — it must be confirmed via execute_refund_order

READ 7 tools

Read discover_manifest Fetch and summarize an AgentBridge manifest from a URL. Use this first to understand what actions a site expos Read get_audit_log Read the local AgentBridge audit log. Filter by manifest URL with the optional Read get_order Returns full details for a single order, including notes and refund history. Read list_actions List all actions in the AgentBridge manifest at a URL. Read list_orders Returns recent orders, optionally filtered by status. Read ping Returns a heartbeat. Useful for connectivity checks. Read scan_agent_readiness Score how agent-ready a URL is. Returns a 0–100 score, structured checks, and grouped recommendations.

EXPLORE

Related servers

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

FAQ

Questions about AgentBridge MCP Server

Can an AI agent move money through the AgentBridge MCP Server MCP server? +

Yes. The AgentBridge MCP Server server exposes 1 financial tools including execute_refund_order. Without a policy, an autonomous agent can call these with no spend caps, no rate limits, and no approval flow. PolicyLayer lets you block financial tools by default, require human approval, or set per-tool rate limits — enforced on every call.

How do I prevent bulk modifications through AgentBridge MCP Server? +

The AgentBridge MCP Server server has 3 write tools including add_internal_note, create_thing, draft_refund_order. 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 AgentBridge MCP Server.

How many tools does the AgentBridge MCP Server MCP server expose? +

12 tools across 4 categories: Execute, Financial, Read, Write. 7 are read-only. 5 can modify, create, or delete data.

How do I enforce a policy on AgentBridge MCP Server? +

Register the AgentBridge MCP Server 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 AgentBridge MCP Server tool call.

Deterministic rules across all 12 AgentBridge MCP Server tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.

ENFORCE POLICY ON AGENTBRIDGE →

Instant setup, no code required.

12 AgentBridge MCP Server tools catalogued and risk-classified — across an index of 43,000+ MCP servers.