Produce exact undo SQL (for DB proposals) or a restore plan (for config).
AI agents use generate_rollback to create or update resources in Safe Migrations — usually the action step of a workflow, after the agent has gathered context. Every call changes real data in your Safe Migrations environment.
generate_rollback produces SQL undo statements and restoration plans, which are reversible data modifications (Write category). It does not execute these changes, so it is not Execute or Destructive. The severity is medium because while the generated SQL could be dangerous if misapplied, the tool itself only produces instructions within a gated migration system.
From the tool's definition Tool description: 'Produce exact undo SQL (for DB proposals) or a restore plan (for config).' The tool generates reversible restoration instructions rather than executing destructive changes. It creates SQL and plans but does not apply them.
Attacks that exploit this kind of access
Produce exact undo SQL (for DB proposals) or a restore plan (for config). It is categorised as a Write tool in the Safe Migrations MCP Server, which means it can create or modify data. Consider rate limits to prevent runaway writes.
Register the Safe Migrations MCP server in PolicyLayer and add a rule for generate_rollback: 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 Safe Migrations. Nothing to install.
generate_rollback 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 generate_rollback 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 generate_rollback. 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.
generate_rollback is provided by the Safe Migrations MCP server (possibly6/safe-migrations-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 →