Skip to content

Latest commit

 

History

History
335 lines (246 loc) · 12 KB

changelog.md

File metadata and controls

335 lines (246 loc) · 12 KB

v0.8.2

v0.8.1

v0.8.0

v0.7.9

  • Thank you merlinfrombelgium
    • Introduce re-prompting after explain is chosen in copilot function
  • Updated to latest models for interacting with GPT

v0.7.8

  • Added example: Read-ExcelAndMultiplyUnits.ps1

Uses the ImportExcel Get-ExcelFileSchema to read the schema of the Excel file, then creates GPT chat messages for the user and system, the prompt read the excel file multiply the units by 20%. Finally, it outputs the PowerShell code that reads the Excel file, multiplies the units by 20%, and formats the output as a table.

v0.7.7

Thank you Shaun Lawrie

v0.7.6

  • Implement mechanism to turn on/off persistence of chat message
  • Add Get-CompletionFromMessages so you can get messages can be passes standalone to get the completions

v0.7.5

  • Fixed spaces in property names #141

v0.7.4

  • Thank you Shaun Lawrie!
    • Fix function builder not running with latest PowerShellAI
    • Fix write-progress breaking notebooks

v0.7.3

  • Thank you Shaun Lawrie!
    • Added Write-Progress, tests and mocks

v0.7.2

  • Added Write-Information as an indicator after callint GPT

v0.7.1

  • New-Chat refactored
    • Now immediately posts to GPT if you specify a prompt and displays the response
    • No prompt, still closes existing session and creates a new one

v0.6.3

  • Improved the system prompt for copilot gh? function
  • Make New-NBCell public

v0.6.2

  • Added Get-OpenAIEmbedding to get a vector representation of a given input
  • Added -temperature param to chat
  • Added "send to vs code" option to menu
  • Added two new functions:
    • git?: Translate natural language to Git commands
    • gh?: Translate natural language to GitHub CLI commands

Full Changelog: https://github.com/dfinke/PowerShellAI/compare/v0.6.1...v0.6.2

v0.6.1

  • Upgraded copilot and explain to use Write-CodeBlock fir syntax highlighting
  • Added call to Invoke-Formatter to prettify the code

v0.6.0

v0.5.8

Thank you @ShaydeNofziger for the contribution!

  • Update .gitignore
  • Update help comment

Also, I:

  • Added kql to validate set for NBCopilot

v0.5.7

Thanks Shaun, great work!

v0.5.6

What's Changed

  • Enables chat conversations with either the public OpenAI or a private Azure OpenAI Service.

Community Contributions:

  • Thank you Svyatoslav Pidgorny
    • Copilot prompt change, adding clipboard

      PS D:\> copilot 'cmds to find k8 pods'
      ╔═════════════════════════╗
      ║Q: cmds to find k8 pods  ║
      ║═════════════════════════║
      ║1: kubectl get pods      ║
      ╚═════════════════════════╝
      Run the code? You can also choose additional actions
      [Y] Yes  [E] Explain  [C] Copy  [N] No  [?] Help (default is "N"):
      

New Contributors

v0.5.5

  • Added support for GPT-4, conversation-in and message-out
    • Saves the conversation to a file in each invocation
    • Supports changing chat options like the model like gpt-4 or gpt-3.5-turbo and more
    • List sessions that have been saved, plus you can get their content

Note: This defaults to using gpt-4. You can set the model to chat with:

Get-ChatSessionOptions
Set-ChatSessionOption -model  gpt-3.5-turbo
Get-ChatSessionOptions

Getting started example:

New-Chat 'respond only in json'
chat 'what are the capitals of Spain, France, and the USA?'
{
  "Spain": "Madrid",
  "France": "Paris",
  "USA": "Washington, D.C."
}

v0.5.4

  • Copilot can now explain the code it generates.
  • Thank you Kris Borowinski for fixing the Pester tests.

v0.5.3

  • Thank you to Matt Cargile for suggestions on improving the prompt for Invoke-AIExplain and reviewing the updates.
    • Prompt now includes
      • 'You are running powershell on'
      • $PSVersionTable.Platform
  • Added $max_tokens to Invoke-AIExplain
  • Added $IdEnd. You can ask for the explanation of a range of history items Invoke-AIExplain -Id 20 -IdEnd 23

v0.5.2

  • Added Invoke-AIExplain - The function utilizes the OpenAI GPT-3 API to offer explanations for the most recently run command, and more.
  • Added alias explain to Invoke-AIExplain

v0.5.1

  • Added proof of concept to work with the new Chat REST API

    • There is more to it. Requires refactoring and tests
    • Proving to be very useful
    new-chat 'you are a powershell bot'
    
    chat 'even numbers btwn 1 and 10'
    chat 'odd numbers'

v0.5.0

  • Thank you Kris Borowinski for re-working Get-OpenAIKey to Get-LocalOpenAIKey and creating/updating tests

v0.4.9

  • For Get-OpenAIUsage
    • Default start and end date
    • Added switch $OnlyLineItems

v0.4.8

Thank you to the community for your contributions!

v0.4.7

Thank you to the community for your contributions!

v0.4.6

  • Thank you to Pieter Jan Geutjens
    • Get-OpenAIEdit works both with pipeline input as well as -InputText param

v0.4.5

  • Added New-Spreadsheet - Creates a new Excel spreadsheet from a prompt
  • Moved [CmdletBinding()] above param. Synopsis was not displaying.
  • Changed the default model for Get-OpenAIEdit to code-davinci-edit-001
  • Thank you Skatterbrainz
  • Thank you Skatterbrainz
  • Thank you Kris Borowinski
    • Wired in A-Z ability to provide OpenAIKey via secure string

v0.4.4

  • Added Get-OpenAIEdit. Given a prompt and an instruction, the model will return an edited version of the prompt. Thank you Skatterbrainz

v0.4.3

  • Added -Method POST to Get-OpenAIModeration. Thank you Skatterbrainz

v0.4.2

  • Change -temperature default to 0

v0.4.1

  • Thank you to Pieter Jan Geutjens
    • Added -temperature param to ai and copilot
    • Changed the input type from int to decimal
    • Changed the range on temperature from [0,1] to [0,2] according to the API documentation

v0.4.0

  • Refactored to use Invoke-OpenAIAPI function. This function is used by all the other functions in the module. This allows for a single place to update the API URL and the API Key.
  • Add Get-* functions for OpenAI URIs
  • Took the function suggestions from Skatterbrainz and updated with Invoke-OpenAIAPI the refactor: https://github.com/dfinke/PowerShellAI/pull/30
  • Refactored Get-DalleImage to use Invoke-OpenAIAPI
  • Refactored Get-GPT3Completion to use Invoke-OpenAIAPI

v0.3.3

  • Check if $result.choices is not null before trying to access it. Thank you StartAutomating
  • Examples added to comment based help in copilot. Thank you Wes Stahler
  • Add New-Spreadsheet script. Creates a new spreadsheet from a prompt. Check out the code
  • Added ConvertFrom-GPTMarkdownTable function. Converts a markdown table to a PowerShell object. Check out the code
  • Unit tests started
  • GitHub Actions in place to run CI/CD

v0.3.2

  • Added Get-DalleImage: Given a description, the model will return an image
  • Added Set-DalleImageAsWallpaper: Given a description, the model will return an image form DALL-E and set it as the wallpaper

v0.3.1

  • Added -max_tokens parameter to the ai function

v0.3.0

  • Added copilot - Makes the request to GPT, parses the response and displays it in a box and then prompts the user to run the code or not. Check the README.md for me details.

  • Added ai function:

    • Experimental function enables piping

      ai "list of planets only names as json"
      [
          "Mercury",
          "Venus",
          "Earth",
          "Mars",
          "Jupiter",
          "Saturn",
          "Uranus",
          "Neptune"
      ]
      ai "list of planets only names as json" |
      ai 'convert to  xml'
      <?xml version="1.0" encoding="UTF-8"?>
      <Planets>
          <Planet>Mercury</Planet>
          <Planet>Venus</Planet>
          <Planet>Earth</Planet>
          <Planet>Mars</Planet>
          <Planet>Jupiter</Planet>
          <Planet>Saturn</Planet>
          <Planet>Uranus</Planet>
          <Planet>Neptune</Planet>
      </Planets>

v0.2.0

  • Thank you Martyn Keigher for your contributions!
    • Added gpt as an alias:

      # Get-GPT3Completion "list of planets only names as json"
      gpt "list of planets only names as json"
      ``
      
    • Added validation for: temperature, max_tokens, top_p, frequency_penalty, presence_penalty