Skip to main content

Build your own slash command

There are two ways to add custom slash commands:

  1. With natural language prompts - this is simpler and only requires writing a string or string template.
  2. With a custom function - this gives you full access to the Continue SDK and allows you to write arbitrary Typescript code.

"Custom Commands" (Use Natural Language)

You can add custom slash commands by adding to the customCommands property in config.json.

  • name: the name of the command, which will be invoked with /name
  • description: a short description of the command, which will appear in the dropdown
  • prompt: a templated prompt to send to the LLM

Custom commands are great when you are frequently reusing a prompt. For example, if you've crafted a great prompt and frequently ask the LLM to check for mistakes in your code, you could add a command like this:

config.json
customCommands=[{
"name": "check",
"description": "Check for mistakes in my code",
"prompt": "{{{ input }}}\n\nPlease read the highlighted code and check for any mistakes. You should look for the following, and be extremely vigilant:\n- Syntax errors\n- Logic errors\n- Security vulnerabilities\n- Performance issues\n- Anything else that looks wrong\n\nOnce you find an error, please explain it as clearly as possible, but without using extra words. For example, instead of saying 'I think there is a syntax error on line 5', you should say 'Syntax error on line 5'. Give your answer as one bullet point per mistake found."
}]

Templating

The prompt property supports templating with Handlebars syntax. You can use the following variables:

  • input (used in the example above): any additional input entered with the slash command. For example, if you type /test only write one test, input will be only write one test. This will also include highlighted code blocks.
  • File names: You can reference any file by providing an absolute path or a path relative to the current working directory.

Custom Slash Commands

If you want to go a step further than writing custom commands with natural language, you can write a custom function that returns the response. This requires using config.ts instead of config.json.

To do this, push a new SlashCommand object to the slashCommands list. This object contains "name", the name that you will type to invoke the slash command, "description", the description seen in the dropdown menu, and "run". The run function is any async generator that should yield strings as you want them to be streamed to the UI. As an argument to the function, you have access to a ContinueSDK object with utilities such as access to certain information/actions within the IDE, the current language model, and a few other utilities. For example, here is a slash command that generates a commit message:

~/.continue/config.ts
export function modifyConfig(config: Config): Config {
config.slashCommands?.push({
name: "commit",
description: "Write a commit message",
run: async function* (sdk) {
const diff = await sdk.ide.getDiff();
for await (const message of sdk.llm.streamComplete(
`${diff}\n\nWrite a commit message for the above changes. Use no more than 20 tokens to give a brief description in the imperative mood (e.g. 'Add feature' not 'Added feature'):`,
{
maxTokens: 20,
},
)) {
yield message;
}
},
});
return config;
}