Mogick
Docs Home

English

简体中文

English

简体中文

Algorithm and script workspace ​

Mogick desktop can put project files, AI sessions and integrated terminals in the same workbench, which is suitable for handling tasks such as script analysis, result checking and documentation organization.

Common scenarios ​

  • Let AI read scripts in your project and interpret inputs, outputs and key parameters.
  • Run commands in the integrated terminal and submit the output to AI for analysis.
  • View AI-generated Markdown, text or data files.
  • Add files that need further processing to the session and let the AI ​​continue working based on the file contents.

Recommendation process ​

  1. Select the item you want to work on on the left.
  2. Describe the algorithm or script task you want to complete in the middle conversation area.
  3. To run a command, open Integrated Terminal.
  4. View the generated or modified files in the Product Workspace on the right.
  5. Continue to ask questions about key results and let AI provide explanations, summaries, or follow-up suggestions.

Example: Let AI read and run a script ​

Taking a data processing script as an example, a typical closed loop is:

  1. Select the project containing the script on the left and send in the conversation area: "Read analysis.jl and explain its input, output and key parameters".
  2. After AI gives an explanation, continue to ask: "Add a piece of code to print a summary of the results on the console" - After AI modifies the file, it will give a [File Editing Card] (/desktop-app/review-workspace), first "View Changes" to confirm the changes.
  3. Open [Integrated Terminal] (/desktop-app/integrated-terminal) to run the script, paste the terminal output back to the session, and let AI continue to analyze or correct based on the actual results.
  4. The generated result files (such as result.csv, summary.md) will appear in the [Product Workspace] (/desktop-app/artifact-workspace) on the right, click to preview.

The entire process does not require switching windows: reading scripts, changing codes, running commands, and viewing products are all completed in the same workbench.

Notes ​

  • Before running commands that will modify files or take a long time, make sure the current project path is correct.
  • Be cautious about external scripts, third-party commands and network requests, and check the meaning of the command first if necessary.
  • If the result file is large, give priority to letting AI summarize the key points before deciding whether to continue in-depth analysis.

FAQ ​

  • **Can AI run scripts directly? ** AI is mainly responsible for reading and modifying code; running commands is performed through Integrated Terminal. You can return the terminal output to the session for AI to continue analysis. When it comes to modifying files or executing commands, confirmation may be requested first (see Permissions and Approval Policies).
  • **What environment is required for the script to run? ** The terminal will execute the command in the current project directory, and the dependent runtime (such as Julia/Python) must be installed on the machine and can be called in the terminal.
  • **Where do the generated files go? ** Files generated or modified by AI will be placed in the project directory and appear in the file tree in the [Product Workspace] (/desktop-app/artifact-workspace) on the right.