help - MadBomber/aia GitHub Wiki

The --help option displays on STDOUT the usage information for the AIA program and then exits.

The help system is the best way to find out which options do what.

aia --help

Will send to STDOUT something that looks like this (Last Update: 2024-01-06)

Usage: aia [options] [PROMPT_ID] [CONTEXT_FILE]*
       aia --chat [PROMPT_ID] [CONTEXT_FILE]*
       aia --chat [CONTEXT_FILE]*
        --chat                       Begin a chat session with the LLM after the initial prompt response; will set --no-out_file so that the LLM response comes to STDOUT.
        --adapter ADAPTER            Interface that adapts AIA to the LLM
    -m, --model MODEL                Name of the LLM model to use
        --terse                      Adds a special instruction to the prompt asking the AI to keep responses short and to the point
    -c, --config_file FILE           Load config file
    -p, --prompts_dir DIR            Directory containing prompt files
        --roles_prefix PREFIX        Subdirectory name for role files (default: roles)
    -r, --role ROLE_ID               Role ID to prepend to prompt
        --refresh DAYS               Refresh models database interval in days
        --regex pattern              Regex pattern to extract parameters from prompt text
    -o, --[no-]out_file [FILE]       Output file (default: temp.md)
    -a, --[no-]append                Append to output file instead of overwriting
    -l, --[no-]log_file [FILE]       Log file
        --md, --[no-]markdown        Format with Markdown
    -n, --next PROMPT_ID             Next prompt to process
        --pipeline PROMPTS           Pipeline of prompts to process
    -f, --fuzzy                      Use fuzzy matching for prompt search
    -d, --debug                      Enable debug output
        --no-debug                   Disable debug output
    -v, --verbose                    Be verbose
        --speak                      Simple implementation. Uses the speech model to convert text to audio, then plays the audio. Fun with --chat. Supports configuration of speech model and voice.
        --voice VOICE                Voice to use for speech
        --sm, --speech_model MODEL   Speech model to use
        --tm, --transcription_model MODEL
                                     Transcription model to use
        --is, --image_size SIZE      Image size for image generation
        --iq, --image_quality QUALITY
                                     Image quality for image generation
        --style, --image_style STYLE Style for image generation
        --system_prompt PROMPT_ID    System prompt ID to use for chat sessions
    -t, --temperature TEMP           Temperature for text generation
        --max_tokens TOKENS          Maximum tokens for text generation
        --top_p VALUE                Top-p sampling value
        --frequency_penalty VALUE    Frequency penalty
        --presence_penalty VALUE     Presence penalty
        --dump FILE                  Dump config to file
        --completion SHELL           Show completion script for bash|zsh|fish - default is nil
        --version                    Show version
    -h, --help                       Prints this help
        --rq, --require LIBS         Ruby libraries to require for Ruby directive
        --tools PATH_LIST            Add a tool(s)
        --at, --allowed_tools TOOLS_LIST
                                     Allow only these tools to be used
        --rt, --rejected_tools TOOLS_LIST
                                     Reject these tools