check_action
Enforce a safety policy on a proposed agent action before execution, blocking unauthorized tools, excessive value, or disallowed recipients.
Instructions
SAFETY: enforce a policy on a proposed agent action (tool call) BEFORE it runs. policy supports forbid_tools/allow_tools, max_value_wei, allow_recipients. Returns {allowed, violations}. Fail-safe: denies on unreadable input. The 'inspect what the agent DOES' gate, not just who it is.
Tool call template
Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.
{
"id": 1,
"jsonrpc": "2.0",
"method": "tools/call",
"params": {
"arguments": {
"action": "<action>"
},
"name": "check_action"
}
}Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| action | Yes | Proposed action: {tool, args}. | |
| policy | No | Rules: forbid_tools[], allow_tools[], max_value_wei, allow_recipients[]. |