// SCAN REPORT

Claude Bridge

16 tools. 8 can modify or destroy data without limits.

3 destructive tools with no built-in limits. Policy required.

Last updated: 2 Jul 2026

8 can modify or destroy data

8 read-only

16 tools total

Community server · catalogue entry verified 02/07/2026

How to control Claude Bridge ↓

TOOL MAP

What Claude Bridge exposes to your agents

Read (8) Write / Execute (5) Destructive / Financial (3)

Running more than one server? Check your whole stack →

MOST DANGEROUS TOOLS

Critical Risk

The most dangerous Claude Bridge tools

Destructive · High cancel_dispatch The description is empty, so classification relies on the tool name alone. Destructive · Medium cancel_schedule Cancelling a schedule is an irreversible action: once stopped, the schedule no longer fires future ticks and there is no indication of an undo/restore mechanism. Destructive · High reset_channel The 'drop' operation permanently removes a channel's pinned session, making this a destructive action. Execute · High wait_any_completion This tool appears to be part of a work-dispatch system that executes code in a devcontainer environment.

8 of Claude Bridge's 16 tools can modify, destroy, or commit something on every call — and an agent calls them with no built-in limits.

POLICY

How to control Claude Bridge

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

Deny destructive operations

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

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

Cap read operations

{
  "bridge_help": {
    "limits": [
      {
        "counter": "bridge_help_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 Claude Bridge — 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 CLAUDE BRIDGE →

Instant setup, no code required.

FULL CATALOGUE

All 16 Claude Bridge tools

DESTRUCTIVE 3 tools

Destructive cancel_dispatch cancel_dispatch Destructive cancel_schedule Stop a schedule from firing further ticks. In-flight ticks are Destructive reset_channel Drop a channel's pinned session so the next dispatch starts a

EXECUTE 5 tools

Execute dispatch dispatch Execute dispatch_async dispatch_async Execute schedule_dispatch schedule_dispatch Execute wait_any_completion wait_any_completion Execute wait_dispatch wait_dispatch

READ 8 tools

Read bridge_help bridge_help Read get_dispatch get_dispatch Read get_schedule Detail view of one schedule. Includes last_job_id so you can Read list_channels channel → pinned session_id. Doesn't tell you whether work is Read list_completions list_completions Read list_events list_events Read list_jobs All tracked jobs (running and finished). raw is stripped from Read list_schedules Every schedule the bridge knows about — active, completed, cancelled, error.

EXPLORE

Related servers

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

FAQ

Questions about Claude Bridge

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

Yes. The Claude Bridge server exposes 3 destructive tools including cancel_dispatch, cancel_schedule, reset_channel. These permanently remove resources with no undo. PolicyLayer blocks destructive tools by default so they never reach the upstream server.

How many tools does the Claude Bridge MCP server expose? +

16 tools across 4 categories: Destructive, Execute, Read, Write. 8 are read-only. 8 can modify, create, or delete data.

How do I enforce a policy on Claude Bridge? +

Register the Claude Bridge 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 Claude Bridge tool call.

Deterministic rules across all 16 Claude Bridge tools. Per-identity grants. Full audit log. Live in minutes. Nothing to install.

ENFORCE POLICY ON CLAUDE BRIDGE →

Instant setup, no code required.

16 Claude Bridge tools catalogued and risk-classified — across an index of 43,000+ MCP servers.