This Document is for Skeet v2.

Skeet AI Mode

Skeet Framework has AI-powered features. Skeet AI enables AI debugging and AI code generation. (At the moment, we recommend using the OpenAI model gpt4.

Skeet AI YouTube - 1/7. Basic Usage (Video)

Skeet AI

There are currently six modes in Skeet AI mode. Each mode uses AI to automate professional tasks.

  • $prisma - Automatic generation of Prisma schema
  • $method - Automatic generation of functions
  • $typedoc - Automatic generation of TypeDoc
  • $firestore - Automatic generation of Firestore models
  • $translate - AI translation
  • $function - Automatic generation of Firebase Functions

Launching Skeet AI Mode

You can launch the AI mode by running the skeet ai command as follows.

bash
$ skeet ai ╔══════════════╤════════════════╗ │ Option │ Value │ ╟──────────────┼────────────────╢ │ Type of AI │ VertexAI │ ╟──────────────┼────────────────╢ │ Model │ chat-bison@001 │ ╟──────────────┼────────────────╢ │ Max Tokens │ 1000╟──────────────┼────────────────╢ │ Temperature │ 0╚══════════════╧════════════════╝ 🤖 Skeet AI Mode $ <mode> to change AI mode 🤖 $ prisma $ typedoc $ translate $ firestore $ function $ method $ help $ q VertexAI has been selected 🤖 (Enter 'q' to quit) ? How can I assist you? You: $ prisma

While Skeet AI is running, you can change the AI mode by entering a command that starts with $.

The default AI is VertexAI. You can also use OpenAI by changing the options.

bash
$ skeet ai --help Usage: skeet ai [options] AI Playground Options: -v, --vertex Vertex AI -o, --openai OpenAI -m, --model <string> Model -token, --token <number> Max Tokens -temp, --temperature <number> Temperature -h, --help display help for command

Setting the Language for Skeet AI

Skeet AI uses the language set in skeet-cloud.config.json. Currently, 17 languages are supported. You can change the language of the AI by modifying the lang property of the ai.

Example of skeet-cloud.config.json:

json
{ // omitted "ai": { "lang": "en", "ais": [ { "name": "VertexAI", "availableModels": ["chat-bison@001", "chat-bison-32k"] } ] } }