AI agents invoke dap_shutdown to trigger actions in Mcp Debugpy. 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.
While termination itself doesn't destructively delete data, it executes a command that forcibly ends a running session/process state. This is closer to Execute (triggers external operations) than Write (reversible data modification) or Destructive (data deletion). However, the severity is high because an AI agent carelessly invoking this could disrupt an active debugging workflow.
From the tool's definition 'Terminate the current DAP adapter session' - this tool terminates a running debugging session, which is an external operation with effects that depend on when and how it is invoked.
Documented attack patterns abuse exactly the kind of access dap_shutdown gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Mcp Debugpy, and nothing reaches the server without passing your rules. This is the rule we recommend for dap_shutdown:
{
"version": "1",
"default": "deny",
"tools": {
"dap_shutdown": {
"limits": [
{
"counter": "dap_shutdown_rate",
"window": "minute",
"max": 10,
"scope": "grant"
}
]
}
}
}dap_shutdown 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.
Free to start. No card required.
Terminate the current DAP adapter session. It is categorised as a Execute tool in the Mcp Debugpy MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the Mcp Debugpy MCP server in PolicyLayer and add a rule for dap_shutdown: 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 Mcp Debugpy. Nothing to install.
dap_shutdown is a Execute tool with high risk. Execute tools should be rate-limited and have argument validation enabled.
Yes. Add a rate_limit block to the dap_shutdown 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.
Set action: deny in the PolicyLayer policy for dap_shutdown. 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.
dap_shutdown is provided by the Mcp Debugpy MCP server (markomanninen/mcp-debugpy). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Mcp Debugpy, add the rest of your stack, and see everything your agents can call. Then put policy on all of it.
Free to start. No card required.
16 Mcp Debugpy tools catalogued and risk-classified — across an index of 43,000+ MCP servers.