Skip to main content

chat

The chat command enables you to interact with OpenAI's GPT in order to get information and remediation suggestions for a specific instance of a vulnerability that was identified in your code.

Notice

Currently supported only for IaC Security vulnerabilities.

Warning

To use this command, the file where the vulnerability was identified must be located in an accessible location.

Data Security

When sending source code to GPT, we protect your sensitive data by anonymizing all passwords and secrets before the content is sent. The query used to identify sensitive data can be seen here.

Usage

./cx chat [flags]

Flags

Glossary
--chat-apikey <string>

Submit the API Key for your Chat GPT account. Follow this link to generate an API key.

--conversation-id <string>

You can submit a conversation ID in order to continue a previous conversation.

--help, -h

Help for the chat command.

--model <string> (Default: gpt-4)

The model of the GPT account that you are using.

--result-file <string>

The full path to the file in which the vulnerability was identified.

--result-line <string>

The line in your code at which the vulnerability was identified.

--result-severity <string>

The severity level of the vulnerability.

--result-vulnerability <string>

The name of the vulnerability that was identified.

--user-input <string>

You can enter text of a question related to the specified vulnerability that you would like to ask GPT. The following are some examples of suggested questions:

  • What is the purpose of this IaC file?

  • What do these results mean?

  • How can I remediate this vulnerability?

Examples