pay-respects/module-request-ai
2025-04-06 20:13:22 +02:00
..
i18n feat: reasoning AI models 2025-04-06 20:13:20 +02:00
src feat: additional prompts 2025-04-06 20:13:22 +02:00
Cargo.toml v0.7.0 2025-04-05 22:31:25 +02:00
README.md feat: additional prompts 2025-04-06 20:13:22 +02:00

Request AI Module

Module for pay-respects to request AI for suggestions.

Compile time features:

  • libcurl (default): Dynamically links to libcurl. Requires OpenSSL when compiling. When disabled, a curl process is spawned instead.

Configurations

Configuration is done via environment variables:

  • _PR_AI_API_KEY: Your own API key
  • _PR_AI_URL: URL used. Any OpenAI compatible URL can be used, e.g.:
    • https://api.openai.com/v1/chat/completions (Note: OpenAI's ChatGPT is very slow)
    • https://api.groq.com/openai/v1/chat/completions
    • http://localhost:11434/api/chat: Ollama
  • _PR_AI_MODEL: Model used
  • _PR_AI_DISABLE: Setting to any value disables AI integration
  • _PR_AI_LOCALE: Locale in which the AI explains the suggestion. Defaults to user system locale
  • _PR_AI_ADDITIONAL_PROMPT: Additional prompts to be included. (Yes, you can include role-playing prompts you pervert)
    • You are a cute catgirl. Always use cute phrases and expressions to prove your cuteness in the note, including cat imitations like nya~, にゃ~, 喵~.

Compile time variables: Default values for the respective variables above when not set

  • _DEF_PR_AI_API_KEY
  • _DEF_PR_AI_URL
  • _DEF_PR_AI_MODEL

If default values were not provided, pay-respects' own values will be used. Your request will be filtered to avoid abuse usages. Request will then be forwarded to a LLM provider that will not use your data for training. This service is provided free and is not guaranteed to always work. Donations would be appreciated:

Donate using Liberapay Donate using Ko-fi
Donate using Stripe Donate using PayPal