AI agents use cache_touch to create or update resources in Prometheus MCP Server — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Prometheus MCP Server environment.
This tool modifies existing data (cache key expiration) in a reversible manner. It is not destructive since expiration time can be adjusted again. It is not read-only since it changes state. The severity is medium because misuse could affect cache behavior and data availability, but the blast radius is limited to TTL modifications rather than data loss or financial impact.
From the tool's definition Tool description states 'Update the expiration time for a key,' which modifies metadata of an existing cache entry without deleting or creating data.
Documented attack patterns abuse exactly the kind of access cache_touch gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Prometheus MCP Server, and nothing reaches the server without passing your rules. This is the rule we recommend for cache_touch:
{
"version": "1",
"default": "deny",
"tools": {
"cache_touch": {
"limits": [
{
"counter": "cache_touch_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}cache_touch 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.
Free to start. No card required.
Update the expiration time for a key. It is categorised as a Write tool in the Prometheus MCP Server MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Prometheus MCP Server MCP server in PolicyLayer and add a rule for cache_touch: 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 Prometheus MCP Server. Nothing to install.
cache_touch 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 cache_touch 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 cache_touch. 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.
cache_touch is provided by the Prometheus MCP Server MCP server (awslabs.prometheus-mcp-server). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Prometheus 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.
805 Prometheus MCP Server tools catalogued and risk-classified — across an index of 43,000+ MCP servers.