Edit a tool, including its source code. Omit properties that you do not want to change.
AI agents use edit_tool to create or update resources in Riza — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Riza environment.
This tool modifies tool definitions and source code but does not permanently delete them or irreversibly destroy data, placing it in the Write category rather than Destructive. However, severity is elevated to 'high' because editing tool source code could enable code injection, privilege escalation, or malicious tool behavior if an AI agent edits tools to perform unintended actions.
From the tool's definition Tool description explicitly states 'Edit a tool, including its source code.' Edit operations modify data (tool definitions and code) reversibly.
Documented attack patterns abuse exactly the kind of access edit_tool gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Riza, and nothing reaches the server without passing your rules. This is the rule we recommend for edit_tool:
{
"version": "1",
"default": "deny",
"tools": {
"edit_tool": {
"limits": [
{
"counter": "edit_tool_rate",
"window": "minute",
"max": 30,
"scope": "grant"
}
]
}
}
}edit_tool 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.
Edit a tool, including its source code. Omit properties that you do not want to change. It is categorised as a Write tool in the Riza MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Riza MCP server in PolicyLayer and add a rule for edit_tool: 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 Riza. Nothing to install.
edit_tool 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 edit_tool 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 edit_tool. 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.
edit_tool is provided by the Riza MCP server (riza-io/riza-mcp). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Riza, 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.
6 Riza tools catalogued and risk-classified — across an index of 43,000+ MCP servers.