Preview refactoring edits across files without applying. Use getCodeActions first.
AI agents call refactorPreview to retrieve information from Patchwork Os without modifying anything — typically the context-gathering step in research, monitoring, and reporting workflows, before the agent takes action elsewhere.
This tool retrieves and displays proposed refactoring changes for user review before any modifications occur. It has no capability to write, execute, delete, or modify actual files. The dependency on 'getCodeActions first' further suggests an informational/analytical role in a refactoring workflow. The preview-only nature makes it purely a Read operation with minimal risk.
From the tool's definition Tool description states 'Preview refactoring edits across files without applying' — the word 'preview' and explicit 'without applying' indicate read-only operation with no side effects.
Documented attack patterns abuse exactly the kind of access refactorPreview gives an agent:
PolicyLayer is an MCP gateway — it sits between your AI agents and Patchwork Os, and nothing reaches the server without passing your rules. This is the rule we recommend for refactorPreview:
{
"version": "1",
"default": "deny",
"tools": {
"refactorPreview": {}
}
} refactorPreview is read-only, so it stays allowed — but everything else on the server is denied unless you say otherwise.
Free to start. No card required.
Preview refactoring edits across files without applying. Use getCodeActions first. It is categorised as a Read tool in the Patchwork Os MCP Server, which means it retrieves data without modifying state.
Register the Patchwork Os MCP server in PolicyLayer and add a rule for refactorPreview: 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 Patchwork Os. Nothing to install.
refactorPreview is a Read tool with low risk. Read-only tools are generally safe to allow by default.
Yes. Add a rate_limit block to the refactorPreview 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 refactorPreview. 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.
refactorPreview is provided by the Patchwork Os MCP server (oolab-labs/patchwork-os). PolicyLayer sits as a proxy in front of this server to enforce policies before tool calls reach the server.
Start from Patchwork Os, 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.
11 Patchwork Os tools catalogued and risk-classified — across an index of 43,000+ MCP servers.