Execute maintenance commands like VACUUM, ANALYZE, or CREATE DATABASE outside of transactions
AI agents invoke execute_maintenance to trigger actions in PostgreSQL Full Access MCP Server. 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.
VACUUM and ANALYZE perform database-wide operations with side effects (garbage collection, index rebuilding, statistics updates). CREATE DATABASE creates new database objects. These are not Read (they modify state), not Write (they're not reversible in the same session), and not Destructive in the strict sense (VACUUM/ANALYZE don't delete user data, though CREATE DATABASE can consume resources).
From the tool's definition Tool executes maintenance commands (VACUUM, ANALYZE, CREATE DATABASE) outside of transactions. These are DDL/DCL operations that trigger external database operations whose effects depend on command arguments.
Documented attack patterns abuse exactly the kind of access execute_maintenance gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and PostgreSQL Full Access MCP Server, and nothing reaches the server without passing your rules. This is the rule we recommend for execute_maintenance:
{
"version": "1",
"default": "deny",
"tools": {
"execute_maintenance": {
"limits": [
{
"counter": "execute_maintenance_rate",
"window": "minute",
"max": 10,
"scope": "grant"
}
]
}
}
}execute_maintenance 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.
Execute maintenance commands like VACUUM, ANALYZE, or CREATE DATABASE outside of transactions. It is categorised as a Execute tool in the PostgreSQL Full Access MCP Server MCP Server, which means it can trigger actions or run processes. Use rate limits and argument validation.
Register the PostgreSQL Full Access MCP Server MCP server in PolicyLayer and add a rule for execute_maintenance: 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 PostgreSQL Full Access MCP Server. Nothing to install.
execute_maintenance 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 execute_maintenance 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 execute_maintenance. 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.
execute_maintenance is provided by the PostgreSQL Full Access MCP Server MCP server (syahiidkamil/mcp-postgres-full-access). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from PostgreSQL Full Access MCP Server, 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.
7 PostgreSQL Full Access MCP Server tools catalogued and risk-classified — across an index of 43,000+ MCP servers.