patch
Patches: pending updates, unattended upgrades, package integrity, kernel audit, CVE lookup
This record as markdown: /tools/defense/patch.md
What patch does on Defense
AI agents use patch to create or update resources in Defense, usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Defense environment.
| Parameter | Type | Required | Description |
|---|---|---|---|
cveId | string | — | CVE ID e.g. CVE-2024-1234 |
action | string | Yes | Patch management action |
dryRun | boolean | — | Preview only |
maxPackages | number | — | Maximum packages to check |
packageName | string | — | Package name to check urgency for |
changed_only | boolean | — | Only show changed files |
package_name | string | — | Specific package to check |
security_only | boolean | — | Only show security-relevant updates |
Parameters from the server's own tool schema.
Why patch is rated Medium
Patching applies security updates to system packages, which is a reversible modification operation (updates can be rolled back). While security-critical and important for hardening, it does not irreversibly destroy data like Destructive operations, nor does it execute arbitrary code (Execute category). The tool's purpose is defensive system maintenance, making it Write rather than Read.
From the tool's definition Tool description indicates 'pending updates, unattended upgrades, package integrity' — these are write operations that modify system software and configurations.
Attacks that exploit this kind of access
The rule that runs patch safely
PolicyLayer is an MCP gateway: it sits between your AI agents and Defense, and checks every tool call against a rule you set before the call runs. Nothing changes on the server itself. For patch, this is the rule to start with:
patch stays usable, but capped: an agent stuck in a loop can't make hundreds of changes a minute. Everything else on the server is denied unless you say otherwise.
The button opens the PolicyLayer dashboard: create your workspace, connect Defense, apply this rule, and every patch call is checked against it from then on.
Questions about patch
Patches: pending updates, unattended upgrades, package integrity, kernel audit, CVE lookup. It is categorised as a Write tool in the Defense MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
patch accepts 8 parameters: cveId, action, dryRun, maxPackages, packageName, changed_only, package_name, security_only. Required: action. The full parameter table on this page comes from the server's own tool schema.
Register the Defense MCP server in PolicyLayer and add a rule for patch: 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 Defense. Nothing to install.
patch is a Write tool with medium risk. Write tools should be rate-limited to prevent accidental bulk modifications.
Yes. Add a rate_limit block to the patch 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 patch. 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.
patch is provided by the Defense MCP server (defense-mcp-server). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
More on Defense, and thousands of servers like it.
This server
Across the catalogue