Apply Multi-File Patches with the apply_patch Tool

Learn how to use the apply_patch tool to apply structured patches across multiple files. This page covers parameters, configuration, and usage notes for developers.

Read this when

  • You need structured file edits across multiple files
  • You want to document or debug patch-based edits

Apply file changes with a structured patch format. This works well for edits spanning multiple files or hunks where a single edit call would be fragile.

The tool expects one input string containing one or more file operations:

*** Begin Patch
*** Add File: path/to/file.txt
+line 1
+line 2
*** Update File: src/app.ts
@@ optional change context
-old line
+new line
*** Delete File: obsolete.txt
*** End Patch

Parameters

  • input (required): The full patch content, including *** Begin Patch and *** End Patch.

Notes

  • Patch paths support both relative paths (relative to the workspace directory) and absolute paths.
  • tools.exec.applyPatch.workspaceOnly defaults to true (workspace-contained). Set it to false only when you intentionally want apply_patch to write or delete outside the workspace directory.
  • Use *** Move to: inside an *** Update File: hunk to rename files.
  • *** End of File marks an EOF-only insert when needed.
  • Enabled by default for every model. Set tools.exec.applyPatch.enabled: false to disable it, or restrict it to specific models with tools.exec.applyPatch.allowModels (accepts raw ids like gpt-5.4 or full ids like openai/gpt-5.4).
  • Configuration is under tools.exec.applyPatch.*.

Example

{
  "tool": "apply_patch",
  "input": "*** Begin Patch\n*** Update File: src/index.ts\n@@\n-const foo = 1\n+const foo = 2\n*** End Patch"
}
  • Diffs, A read-only diff viewer for presenting changes.

  • Exec tool, Runs shell commands from the agent.

  • Code execution, Sandboxed remote Python analysis with xAI.