Back to Goose

Code Mode

documentation/docs/guides/managing-tools/code-mode.md

1.33.13.6 KB
Original Source

import Tabs from '@theme/Tabs'; import TabItem from '@theme/TabItem';

Code Mode is a method of interacting with MCP tools programmatically instead of calling them directly. Code Mode is particularly useful when working with many enabled extensions, as it can help manage context window usage more efficiently.

:::info This functionality requires the built-in Code Mode extension to be enabled. :::

Code Mode controls how tools are discovered and called:

  • Tools from enabled extensions are discovered on-demand and loaded into context as needed
  • Multiple tool calls are batched in one execution
  • Intermediate results are chained (output from one tool as input to the next)

How Code Mode Works

The Code Mode extension is an MCP server that uses the MCP protocol to expose three foundational meta-tools. When Code Mode is enabled, goose switches to Code Mode. For every request, the LLM writes JavaScript code that goose executes using pctx (Port of Context) (GitHub), a custom Deno-based runtime, to:

  • Discover available tools from your enabled extensions (if needed)
  • Learn how to work with the tools it needs for the current task
  • Call those tools programmatically to complete the task

Traditional vs. Code Mode Tool Calling

Traditional MCP tool calling and Code Mode are two different approaches to the same goal: giving goose access to tools.

AspectTraditionalCode Mode
Tool DiscoveryAll tools from enabled extensions, for example:
developer.shell
developer.text_editor
github.list_issues
github.get_pull_request
slack.send_message
• ... potentially many moreCode Mode extension's meta-tools:
list_functions
get_function_details
execute_typescript

The LLM uses these tools to discover tools from other enabled extensions as needed | | Tool Calling | • Sequential tool calls • Each result sent to the LLM before the next call | • May require tool discovery calls • Multiple tool calls batched in one execution • Intermediate results are chained and processed locally | | Context Window | Every LLM call includes all tool definitions from enabled extensions | Every LLM call includes the 3 meta-tool definitions, plus any tool definitions previously discovered in the session | | Best For | • 1-3 enabled extensions • Simple tasks using 1-2 tools | • 5+ extensions • Well-defined multi-step workflows |

:::info Text-Only Results Code Mode only supports text content from tool results. Images, binary data, and other content types are ignored. :::

Additional Resources

import ContentCardCarousel from '@site/src/components/ContentCardCarousel'; import gooseCodeMode from '@site/blog/2025-12-15-code-mode-mcp/header-image.jpg'; import notMcpReplacement from '@site/blog/2025-12-21-code-mode-doesnt-replace-mcp/header-image.png';

<ContentCardCarousel items={[ { type: 'blog', title: 'Code Mode for MCP', description: 'Learn about the code execution approach to MCP tool calling.', thumbnailUrl: gooseCodeMode, linkUrl: '/blog/2025/12/15/code-mode-mcp', date: '2025-12-15', duration: '5 min read' }, { type: 'blog', title: 'Code Mode Doesn't Replace MCP', description: 'Understanding how Code Mode and MCP work together.', thumbnailUrl: notMcpReplacement, linkUrl: '/blog/2025/12/21/code-mode-doesnt-replace-mcp', date: '2025-12-21', duration: '8 min read' } ]} />