A powerful typer based Deepseek AI Assistant you can use to engineer TODAY.
cp .env.sample .env
- Update with your keys
uv sync
- (optional) install python 3.11 (
uv python install 3.11
)
- 🧠 Brain:
Deepseek V3
- 📝 Job (Prompt(s)):
prompts/typer-commands.xml
- 💻 Active Memory (Dynamic Variables):
scratchpad.txt
+commands/template.py
- 👂 Ears (STT):
RealtimeSTT
- 🎤 Mouth (TTS):
ElevenLabs (Flash 2.5)
- 🧠 Brain:
Deepseek V3
- 🧠 Brain:
Deepseek V3
- 📝 Job (Prompt(s)):
None
- 💻 Active Memory (Dynamic Variables):
scratchpad.txt
+commands/template.py
- 👂 Ears (STT):
RealtimeSTT
- 🎤 Mouth (TTS):
RealtimeTTS
Start a conversational chat session with the base assistant:
python main_base_assistant.py chat
This will start an interactive chat session where you can:
- Type messages to the assistant
- Get spoken responses using TTS
- Type 'exit' or 'quit' to end the session
- Use Ctrl+C to gracefully exit
The base assistant features:
- 🧠 Brain: Deepseek V3
- 🎤 Mouth: System TTS (can be replaced with Azure/11Labs)
- 📝 Memory: Maintains full conversation history
- 📊 Logging: Detailed session logs in output/ directory
# in a python env
python main_typer_assistant.py awaken --typer-file commands/template.py --scratchpad scratchpad.md --mode execute
# raw
uv run python main_typer_assistant.py awaken --typer-file commands/template.py --scratchpad scratchpad.md --mode execute
These commands mirror a single text command to the Deepseek AI Assistant, execute it and returns the output of the command.
# Basic commands
python main.py deep --typer-file commands/template.py --scratchpad scratchpad.txt --prompt "Ping the server"
python main.py deep --typer-file commands/template.py --scratchpad scratchpad.txt --prompt "Ping the server be sure to wait"
python main.py deep --typer-file commands/template.py --scratchpad scratchpad.txt --prompt "show config verbose"
# User management
python main.py deep --typer-file commands/template.py --scratchpad scratchpad.txt --prompt "new user for tim role is admin"
python main.py deep --typer-file commands/template.py --scratchpad scratchpad.txt --prompt "delete user 12345 --confirm"
# File operations
python main.py deep --typer-file commands/template.py --scratchpad scratchpad.txt --prompt "backup data from db/ dir"
python main.py deep --typer-file commands/template.py --scratchpad scratchpad.txt --prompt "download file from http://test.com output to sessions dir retry 8 times"
python main.py deep --typer-file commands/template.py --scratchpad scratchpad.txt --prompt "upload changes to our source_dir"
# Logs and debugging
python main.py deep --typer-file commands/template.py --scratchpad scratchpad.txt --prompt "search logs for 'focus cat off re john' case sensitive"
python main.py deep --typer-file commands/template.py --scratchpad scratchpad.txt --prompt "summarize logs from /var/logs lines=500"
# Data operations
python main.py deep --typer-file commands/template.py --scratchpad scratchpad.txt --prompt "filter records from users.csv query='active=true' limit=100"
python main.py deep --typer-file commands/template.py --scratchpad scratchpad.txt --prompt "transform data input.csv format=json columns=name,email"
# System operations
python main.py deep --typer-file commands/template.py --scratchpad scratchpad.txt --prompt "health check auth_service timeout=60 alert"
python main.py deep --typer-file commands/template.py --scratchpad scratchpad.txt --prompt "sync remotes production --force"
- add arbitrary 'scratchpad' files (make it a list)
- add ears (STT)
- add minimal mouth (TTS) to summarize what was done (no need for brain here just a simple conversational, short summary)
- [] add a minimal barebones stt, brain, tts personal ai assistant.
- [defer] add previous command map history. 1. 'deep go ahead and backup the db' -> 'backup-data db/'\n2. '...'
- this lets us say refer to previous commands.
- LOCAL SPEECH TO TEXT: https://github.com/KoljaB/RealtimeSTT
- faster whisper (support for RealtimeSTT) https://github.com/SYSTRAN/faster-whisper
- examples https://github.com/KoljaB/RealtimeSTT/blob/master/tests/realtimestt_speechendpoint_binary_classified.py
- elevenlabs voice models: https://elevenlabs.io/docs/developer-guides/models#older-models