V1.2.0

Version 1.2.0: What's new

• Introduced an authentication token mechanism to safeguard against server misuse.
• Fixed an issue relating to lost connections without recovery attempts.
• Added logging capability for improved traceability and debugging.
• Migrated modifiable variables to a dedicated configuration file to enhance customization.
• Transformed the server application to run as a Windows service for seamless execution.
• Digitally signed the setup to eliminate antivirus warnings.

v1.1.0

ChatGPT-Executor V1.1.0

News

• Optimized maximum word usage.
• Improved the Executor's behavior when requesting continuation from ChatGPT. It now specifies the last received line and instructs ChatGPT not to resend previously received content.
• Enhanced the management of the initial prompt by moving it to the server side. This reduces dependency on the plugin protocol and ensures ongoing compatibility between different versions. To enable this functionality, a new config parameter has been added on the plugin side. When new communication is detected, the plugin can now send a command to the server using the "_START_NEW" protocol-independent command.

For more information:

• Find the README
• See a Demo youtu.be/9fCtMJQxQ4c
• Check the ChatGPT-Bridge release from here

ChatGPT-Executor_V1.0.0-rc1

ChatGPT-Executor_V1.0.0-rc1

This version is intended to work with ChatGPT-Bridge_V1.0.0-rc1. this version :

• Supports receiving multi commands in the same request, execute them and concatenate answers in one prompt.
• When a output of executed command is bigger than 2000 word ( essentially when GPT ask to read files) will manage to divide it into blocks, and manage to send it in consecutive frames ( to manage the limitation of one prompt size).
• Detect when a received command is incomplete and ask for continuation ( resolve the problem of GPT hanging).

How to use:

• unzip the package the ChatGPT-Executor_V1.0.0-rc1-rc.zip
• launch ChatGPT-Executor.exe