AI agents use set_pyfilter_code to create or update resources in Firegex — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Firegex environment.
This tool modifies (replaces) Python filter code for a Firegex service. While reversible (another set_pyfilter_code call could restore it), it directly alters the security filtering logic. The high severity reflects that corrupted or malicious filter code could compromise the firewall's ability to protect the system, affecting all traffic through that service.
From the tool's definition Tool name 'set_pyfilter_code' indicates replacement of source code. Description states 'Replace the Python filter source for a service.' This is a write operation that modifies code configuration for an active security service.
Attacks that exploit this kind of access
Replace the Python filter source for a service. It is categorised as a Write tool in the Firegex MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Firegex MCP server in PolicyLayer and add a rule for set_pyfilter_code: 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 Firegex. Nothing to install.
set_pyfilter_code 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 set_pyfilter_code 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 set_pyfilter_code. 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.
set_pyfilter_code is provided by the Firegex MCP server (umbra2728/firegex-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Every MCP server has a record like this.
Type a name, get the same breakdown: verified identity, auth posture, risk grade, capabilities, recommended policy.
Teams ship this data inside their own products. See what a licence covers →